Electricity Market Reform An International Perspective

Electricity Market Reform: An International Perspective

Elsevier Internet Homepage-http://www.elsevier.com Consult the Elsevier homepage for full catalogue information on all books, journals and electronic products and services Elsevier Titles of Related Interest DICTIONARY OF ENERGY Cleveland & Morris 2005; ISBN: 0-08-044578-0 www.elsevier.com/locate/doe AGILE ENERGY SYSTEMS Clark & Bradshaw 2004; ISBN: 0-08-044448-2 www.elsevier.com/locate/isbn/0080444482 RESHAPING EUROPEAN GAS AND ELECTRICITY INDUSTRIES Finon & Midttun 2004; ISBN: 0-08-044550-0 www.elsevier.com/locate/isbn/0080445500 Related Journals The following titles can all be found at: http://www.sciencedirect.com The Electricity Journal Applied Energy Climate Policy Energy Energy Conversion & Management Energy Policy Renewable Energy Renewable & Sustainable Energy Reviews Solar Energy Progress in Nuclear Energy Refocus To Contact the Publisher Elsevier welcomes enquiries concerning publishing proposals: books, journal special issues, conference proceedings, etc. All formats and media can be considered. Should you have a publishing proposal you wish to discuss, please contact, without obligation, the Publisher responsible for Elsevier’s energy publishing program: Henri van Dorssen Publisher, Energy Elsevier Limited The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK Tel.: ⫹ 44 1865 84 3931 Fax: ⫹ 44 1865 84 3920 E-mail: [email protected] General enquiries including placing orders, should be directed to Elsevier’s Regional Sales Offices – please access the Elsevier internet homepage for full contact details.

Electricity Market Reform: An International Perspective Edited by Fereidoon P. Sioshansi and Wolfgang Pfaffenberger

AMSTERDAM – BOSTON – HEIDELBERG – LONDON – NEW YORK – OXFORD PARIS – SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO

ELSEVIER B.V. Radarweg 29 P.O. Box 211, 1000 AE Amsterdam The Netherlands

ELSEVIER Inc. 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA

ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK

ELSEVIER Ltd 84 Theobald’s Road London WC1X 8RR UK

© 2006 Elsevier Ltd. All rights reserved. This work is protected under copyright by Elsevier Ltd., and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK: phone (⫹44) 1865 843830, fax (⫹44) 1865 853333, e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone (⫹1) (978) 7508400, fax: (⫹1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (⫹44) 20 7631 5555; fax: (⫹44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2006 ISBN-13: 978-0-08045-030-8 ISBN-10: 0-080-45030-X Printed in Great Britain. 06 07 08 09 10

10 9 8 7 6 5 4 3 2 1

Contents Contributors Foreword: The Market Versus Regulation Stephen Littlechild Introduction to Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies Paul L. Joskow PART I: What’s Wrong With the Status Quo? 1.

Why Restructure Electricity Markets? Fereidoon P. Sioshansi and Wolfgang Pfaffenberger

2.

Sector-Specific Market Power Regulation versus General Competition Law: Criteria for Judging Competitive versus Regulated Markets Günter Knieps

PART II: Trailblazers

vii xvii 1

33 35

49

75

3.

Chile: Where It All Started Ricardo Raineri

4.

Electricity Liberalization in Britain And The Evolution of Market Design David Newbery

109

5.

The Nordic Electricity Market: Robust By Design? Eirik S. Amundsen, Lars Bergman and Nils-Henrik M. von der Fehr

145

PART III: Evolving markets 6.

The Electricity Industry in Australia: Problems Along the Way to a National Electricity Market Alan Moran

77

171

173

7.

Restructuring the New Zealand Electricity Sector 1984–2005 Geoff Bertram

203

8.

Energy Policy and Investment in the German Power Market Gert Brunekreeft and Dierk Bauknecht

235

9.

Competition in the Continental European Electricity Market: Despair or Work in Progress? Reinhard Haas, Jean-Michel Glachant, Nenad Keseric and Yannick Perez

265

v

vi

Contents

PART IV: North America, New world, New Challenges

317

10.

California Electricity Restructuring, The Crisis, and Its Aftermath James L. Sweeney

319

11.

Texas: The Most Robust Competitive Market in North America Parviz Adib and Jay Zarnikau

383

12.

Electricity Restructuring in Canada Michael J. Trebilcock and Roy Hrab

419

13.

The PJM Market Joseph Bowring

451

14.

Independent System Operators in The USA: History, Lessons Learned, and Prospects Richard O’Neill, Udi Helman, Benjamin F. Hobbs and Ross Baldick

15.

Competitive Retail Power Markets and Default Service: The US Experience Taff Tschamler

479 529

PART V: Other Markets

563

16.

The Case of Brazil: Reform By Trial And Error? João Lizardo R. Hermes de Araújo

565

17.

Understanding The Argentinean and Colombian Electricity Markets Isaac Dyner, Santiago Arango and Erik R. Larsen

595

18.

A New Stage of Electricity Liberalization in Japan: Issues and Expectations Mika Goto and Masayuki Yajima

617

Index

645

Contributors

Parviz ADIB is Director of Wholesale Market Oversight at the Public Utility Commission of Texas where he is engaged in designing competitive electricity market rules, providing policy recommendations to the Commissioners, and supervising wholesale market monitoring functions. Prior to joining the Commission, he taught graduate and undergraduate courses in The University of Texas at Austin. He, along with a couple of other Commission Staff, was the recipient of Center for the Advancement of Energy Markets’ Unsung Heroes Award for leadership by a civil servant who has made a significant difference in gas and electric competition in 2005. Dr. Adib’s main research interests include energy and public policy, efficient operation of restructured electricity markets, and effective market monitoring mechanism. He has published papers and prepared numerous presentations and policy recommendations on these topics, notably covering Restructured Electricity Markets. Dr. Adib completed his BA and MS studies in Economics from Tehran University and holds a PhD from The University of Texas at Austin. Eirik S. AMUNDSEN is Professor of Economics at University of Bergen, Norway and at Institute of Food and Resource Economics, The Royal Veterinary and Agricultural University, Copen hagen, Denmark. He has served as Nordic Research Professor affiliated with The Nordic Energy Research Programme, and as a Scientific Adviser to SNF, Norway and SNS, Sweden. He has contributed a variety of articles in theoretical and applied economic journals related to energy economics, environmental economics, and resource economics. In the last few years his research has focused on electricity markets. He has his university degrees from University of Bergen, University of Copenhagen, Institut Français du Pétrole (ENSPM) and Stanford University. He is Dr.ès.sc.écon. (dr. d’Etat) with a Prize-awarded thesis from University of Paris, Panthéon-Assas. Santiago ARANGO is at the Universidad Nacional de Colombia and is about to finish the PhD in the System Dynamics Group, University of Bergen, Norway. He is exploring the dynamics of deregulated electricity markets with experimental economics and system dynamics. From 1998 until 2002 he has done consulting work for governmental and private organization in Colombia, attached to the Energy Institute, Universidad Nacional de Colombia. João Lizardo R.H. de ARAÚJO is Professor of Energy Economics at the Institute of Economics, Federal University of Rio de Janeiro, presently loaned to Eletrobrás to head the Centre for Electric Power Research (CEPEL). Most of his 40-year academic career has been at UFRJ, in the Coordination for Graduate Programmes in Engineering – COPPE (1970–1994) and in the Institute of Economics (1994-present). He has taken a number of positions, including Head of Computing Department at the Institute of Mathematics, Head of Systems and Optimization Programme and of the Interdisciplinary Energy Programme, both at COPPE, and Research Director at the Institute of Economics. Prof. Araújo has also been a Visiting Scholar at the Imperial College, at Lawrence Berkeley Laboratory, and at SPRU. vii

viii

Contributors

Prof. Araújo’s main fields of interest include system optimization, modeling and regulatory issues. He has published numerous papers on energy regulation and energy modeling, as well as on technological innovation. His latest book on energy regulation issues is “Diálogos da Energia: Reflexões sobre a última década 1994–2004” (7 Letras, Rio de Janeiro 2005), with Prof. Oliveira, also of IE/UFRJ. Other relevant publications include “Brazilian Energy Policy: Changing Course?” (with A. de Oliveira). Prof. Araújo has a Degree in Electronic Engineering from Instituto Tecnológico de Aeronáutica, a DEA and a Dr. Sp. from Université de Toulouse. Ross BALDICK is Professor of Electrical and Computer Engineering at The University of Texas at Austin. From 1991 to 1992 he was a Post-Doctoral Fellow at the Lawrence Berkeley Laboratory. In 1992 and 1993 he was an Assistant Professor at Worcester Polytechnic Institute. Dr. Baldick has published over 40 refereed journal articles and has research interests in a number of areas in electric power. His current research involves optimization and economic theory applied to electric power system operations, the public policy and technical issues associated with electric transmission under deregulation, and the robustness of the electricity system to terrorist interdiction. In 1994, Dr. Baldick received a National Science Foundation Young Investigator Award. He is Editor of IEEE Transactions on Power Systems and the Chairman of the System Economics Sub-committee of the Power Systems Analysis, Computation, and Economics Committee of the IEEE Power Engineering Society. He received his BSc and BE Degrees from the University of Sydney, Australia and his MS and PhD from the University of California, Berkeley. Lars BERGMAN is Professor of Economics and President of the Stockholm School of Economics. He is also a Member of the Royal Swedish Academy of Engineering Sciences and Chairman of the Swedish Association for Energy Economics. His research has been focused on general equilibrium modeling, environmental economics and policy, and the economics of electricity markets. During more than a decade he has had an active role in the management of both the Nordic Energy Research Program and an Energy Market Research Program at the Stockholm School of Economics. He is one of the authors of A European Market for Electricity (1999). He has a PhD in Economics and an MSc from the Stockholm School of Economics. Geoff BERTRAM is Senior Lecturer in Economics at Victoria University of Wellington, New Zealand. His primary research interests are in the economics of regulation and industry restructuring, with particular reference to energy and infrastructure sectors; and the economics of development in small island states. Prior to joining the faculty of Victoria University he was employed as a Research Officer at the Institute of Economics and Statistics at Oxford University, writing an economic history of Peru. Dr. Bertram has published numerous papers on energy sector restructuring, including most recently “Price–Cost Margins and Profit Rates in New Zealand Electricity Distribution Networks Since 1994: The Cost of Light-Handed Regulation”, Journal of Regulatory Economics 27(3): 281–307, May 2005; and “Deregulation and Monopoly Profits in New Zealand’s Gas and Electricity Sectors”, Energy Studies Review 12(3): 208–227, Spring 2004. Dr Bertram holds a BA Honors Degree from Victoria University of Wellington, and an MPhil and DPhil from the University of Oxford, UK. He is on the editorial advisory boards of a number of journals, including Environment and Development Economics and Asia Pacific Viewpoint, and was formerly on the editorial board of World Development.

Contributors

ix

Gert BRUNEKREEFT is a Researcher at the Tilburg Law and Economics Center (TILEC) at Tilburg University in the Netherlands. Before that he held research positions in Applied Economics at the University of Cambridge, where he was part of the Cambridge-MIT-Institute (CMI) Electricity Project and before that at the Institute of Transport Economics at Freiburg University in Germany. Dr. Brunekreeft’s main research interests are in industrial economics, regulation, and competition policy of network industries, especially electricity markets. His book Regulation and Competition Policy in the Electricity Market published in 2003 analyses developments on the German electricity market. He has published widely in a variety of energy-related academic journals, including Utilities Policy, Oxford Review of Economic Policy, and Energy Journal. Dr. Brunekreeft holds a Degree in Economics from the University of Groningen in the Netherlands and a PhD and Habilitation from Freiburg University in Germany, both in economics. At the moment he is Associate Editor for the Journal of Network Industries. Dierk BAUKNECHT is Research Fellow with the Oeko-Institut, a German research institute in the area of applied environmental and energy research. He currently works on power plant investment models, network regulation, decentralized power generation, innovation research and transformation of energy systems. Previously, he was market analyst for Germany and modeling manager with a UK-based power market consultancy. His main research interests include the integration of distributed generation into power networks and markets. He has published a number of papers both on this topic and on the development of competition in the German electricity market, mainly in industry journals and the Financial Times. Bauknecht graduated in Political Science at the Free University of Berlin and holds an MSc in Science and Technology Policy from SPRU at the University of Sussex, UK. He is currently completing his PhD on Network Regulation and Distributed Generation. Isaac DYNER is Professor of Operational Research and Energy at the Energy Institute, Universidad Nacional de Colombia. He has been a Visiting Professor of the British Academy and Academic Visitor at Imperial College London, City University, London Business School, Warwick University and several Latin American universities. He has been Editor of the Energy Journal Energetica and Member of the Colombian Council for Energy Research. He has consulted with companies and governments in the areas of strategy, deregulation, energy, and modeling. He has published over 150 research papers in the areas of energy, operational research, and system dynamics. He has published in scientific journals including, among others, Journal of Operational Research, Energy Policy, Utilities Policy, International Journal of Global Energy Issues, Futures, International Journal of Operational Research and System Dynamics Review as well as three books and two chapter in books published by Wiley and Risk Books. He obtained his PhD from London Business School and Master of Sciences from both University of Warwick and Southampton University, UK. Nils-Henrik VON DER FEHR is Professor of Economics at the University of Oslo, where is engaged in research in the fields of competition policy, regulation and energy and environmental economics. Prof. von der Fehr’s main research interests include the design of electricity markets, competitive wholesale auctions, and investment incentives. He has published numerous papers on such topics. He has acted as advisor on regulatory matters to governments around the world, including Australia, Brazil, the Netherlands, Norway, and Sweden. He

x

Contributors

was a Member of the Dutch Electricity Market Surveillance Committee and recently chaired a commission on the developments of the Nordic electricity market. He has a Degree in Economics and Mathematics from the University of Oslo and holds a PhD in Economics from the same university. Jean-Michel GLACHANT is Professor of Economics at University Paris XI, Head of its Economics Department, Member of the Board of the International Society for New Institutional Economics (ISNIE), and Director of the New Institutional Economics Research Center at La Sorbonne University 1991–2000. Since 1995 he has managed a series of research projects on utility privatization and regulation, electricity markets competition, and restructuring. He his or has been Advisor for the French energy regulator, the French government, the European Commission (Energy or Competition), and coordinator of the EU-funded project SESSA on European electricity reforms. He is Founding Member of the new “European Energy Institute”. The two most recent books he edited are Competition in European Electricity Markets: A Cross-Country Comparison, Edward Elgar, 2003 and The Economics of Contracts: Theories and Applications, Cambridge University Press, 2002. He has a Degree in Economics from La Sorbonne University in Paris and holds a PhD in Economics from the same university. Mika GOTO is Research Economist at Socio-Economic Research Center, Central Research Institute of Electric Power Industry (CRIEPI). She was a Visiting Researcher at the Institute of Energy Economics at the University of Cologne and the National Regulatory Research Institute at the Ohio State University. Her main areas of interest include structural reforms, productivity and efficiency analysis as well as cost structure of electric power utilities. She is engaged in research of the economies of scale and productivity growth of electric power industry. Dr. Goto obtained her Doctoral Degree in Economics from Nagoya University, Japan. Reinhard HAAS is Associate Professor of Energy Economics at Vienna University of Technology in Austria. Since 2001 he has served as the Vice-Head of the Institute of Power Systems and Energy Economics. He has been engaged in research projects for the European Commission and others. His current research includes dissemination strategies for renewables; sustainable energy systems, and liberalization versus regulation of energy markets. He has published numerous articles in peer-reviewed international journals. He has studied industrial economics in mechanical engineering. He holds a PhD in Energy Economics from Vienna University of Technology. Udi HELMAN is an Economist at the Federal Energy Regulatory Commission (FERC). He has worked extensively on US electricity market design issues, including auction markets for wholesale energy, ancillary services and installed capacity, transmission usage pricing and market rules for transmission property rights. His major projects have included the initial design and redesign of the Independent System Operator (ISO) New England markets, the Standard Market Design initiative and on the development of the Midwest ISO markets. Dr. Helman has authored or co-authored a number of articles and book chapters on aspects of electricity regulatory reform. These include “Market power monitoring and mitigation in the US wholesale power markets”, Energy (May–June 2006); “A Primer on ComplementarityBased Equilibrium Modeling for Electric Power Markets”, in D.W. Bunn (ed.), Modeling Prices in Competitive Electricity Markets (Wiley, 2004); and “A Joint Energy and Transmission Rights Auction: Proposal and Properties”, IEEE Transactions on Power Systems (November 2002).

Contributors

xi

Dr. Helman holds BA and MA Degrees from the University of Toronto and a PhD from The Johns Hopkins University in Energy Economics and Systems Analysis. Benjamin F. HOBBS is Professor in the Department of Geography and Environmental Engineering of The Johns Hopkins University. He also holds a joint appointment with the Department of Applied Mathematics and Statistics. Previously, he was a Professor of Systems Engineering and Civil Engineering at Case Western Reserve University. He has also been a Member of the Research Staffs of Brookhaven and Oak Ridge National Laboratories. Dr. Hobbs’ research interests are in electric sector policy and planning, and ecosystem management. He is Member of the California ISO Market Surveillance Committee, Scientific Advisor to the ENC Policy Studies Unit, and Member of the Public Interest Advisory Committee of the Gas Technology Institute. His recent books are Energy Decisions and the Environment, A Guide to the Use of Multicriteria Methods (with P. Meier, Kluwer, 2000) and The Next Generation of Electric Power Unit Commitment Models (edited with M. Rothkopf, R. O’Neill, and H. Chao, Kluwer, 2001). His PhD is in Environmental System Engineering from Cornell University. Roy HRAB is Policy Advisor at the Ontario Energy Board where he is engaged in regulatory policy development. Prior to joining the Board he worked at the Institute for Competitiveness and Prosperity and for the Government of Ontario’s Panel on the Role of Government. Mr. Hrab’s current research interests include study of restructured electricity markets and transmission planning. He has published papers and articles (with Michael J. Trebilcock) on the electricity restructuring experience in Ontario, Canada, notably “Electricity Restructuring in Ontario”, in The Energy Journal, 2005. Roy has earned BA and MA Degrees in Economics from the University of Toronto. Paul L. JOSKOW is Elizabeth and James Killian Professor of Economics and Management at MIT and Director of the MIT Center for Energy and Environmental Policy Research. He received his PhD in Economics from Yale University in 1972 and has been on the MIT faculty since then. At MIT he is engaged in teaching and research in the areas of industrial organization, energy and environmental economics, and government regulation of industry. Prof. Joskow has published six books and over 100 papers on topics in these areas. He began doing research and writing on competitive electricity markets over 20 years ago and was co-author (with Richard Schmalensee) of Markets for Power published by MIT Press in 1983. He is Fellow of the Econometric Society and a Fellow of the American Academy of Arts and Sciences. Nenad KESERIC holds a Degree in Electrical Engineering (Power Engineering and Electrical Energetic Systems) at Cacak College of Engineering, University of Technology Kragujevac, Serbia and Montenegro. He is working on a PhD at the Energy Economics Group at Vienna University of Technology. His major focus of work is cross-border electricity trading, congestion management, energy modeling, and price forecasting. He has consulted for regulatory commission and private companies. Günter KNIEPS is Professor of Economics at the University of Freiburg, Germany, where he is engaged in research in the field of network economics. Prior to joining the Faculty of

xii

Contributors

Economics in Freiburg he held a Chair of Microeconomics at the University of Groningen, the Netherlands. He is Member of the Scientific Council of the Ministry of Economics and Labor as well as of the Ministry of Transportation and Housing. Dr. Knieps’ main research interests include study of network economics, de-regulation, competition policy, industrial economics, sector studies on telecommunications, transportation, and energy. He has published widely in academic and professional journals. Dr. Knieps has diplomas in Economics and Mathematics from the University of Bonn, Germany, and a PhD in Mathematical Economics from the University of Bonn. Erik LARSEN is Professor of Management, University of Italian Switzerland, Lugano, Switzerland. Erik previously held appointments at Cass Business School, University of Bologna, London Business School and Copenhagen Business School. During the period 1996–1998 he was an EU Marie Curie Fellow at the University of Bologna. His teaching and research is in the areas of strategy and deregulation of utility companies. His research has been published in journals such as American Metric Journal, European Journal of Operational Research, Energy Policy, Utility Policy, and Management Science. He has consulted for private companies, international organizations and governments in the area of deregulation. He obtained his PhD from the Institute of Economics, Copenhagen Business School and his MSc from the Technical University of Denmark. Stephen LITTLECHILD is Emeritus Professor at the University of Birmingham and Senior Research Associate at the Judge Business School, University of Cambridge. He has been a Member of Ofgem’s Panel of Economic Advisers since 1999. He was the UK’s Director-General of Electricity Supply and Head of the Office of Electricity Regulation (OFFER) from 1989 to 1998. Previously he was Professor of Commerce and Head of Department of Industrial Economics and Business Studies at the University of Birmingham from 1975 to 1989. He was a Member of the Monopolies and Mergers Commission from 1983 to 1989. Prof. Littlechild is an international consultant on privatization, regulation, and competition, especially in the electricity and telecommunications sectors. Since 1999, he has been a policy advisor to governments, regulators, and companies in Chile, Mexico, Philippines, India, Australia, New Zealand, Thailand, Brazil, Poland, Romania, Canada, Saudi Arabia, Russia, China, Argentina, and Colombia, as well as the UK. His recent research interests in electricity include competition in retail supply, transmission regulation, and negotiated settlements. He has a Bachelor of Commerce Degree from the University of Birmingham, a PhD from the University of Texas at Austin, and Honorary Doctorate Degrees from the Universities of Birmingham and East Anglia. Alan MORAN is the Director of the Deregulation Unit at the Institute of Public Affairs in Melbourne, Australia. He was previously a Senior Official in the Australian Federal and Victorian Governments responsible for regulatory review and energy policy matters. He is one of Australia’s best-known commentators on the energy industry and policy having published over 30 major papers covering specific aspects of the industry. He has authored four books, three on environmental economics, and published dozens articles and submissions on privatization, energy, and other economic policy matters. He has Degrees in Economics from the London School of Economics, Management from the University of Salford and a PhD from the University of Liverpool.

Contributors

xiii

David NEWBERY is Professor of Applied Economics at the University of Cambridge, where he is Research Director of the Electricity Policy Research Group. He has been an academic at the Faculty of Economics since 1966, with sabbaticals at Yale, Princeton, Stanford and Berkeley, as well as a spell as Division Chief at the World Bank. He has been an Economic Advisor to Ofgem, Ofwat, and a Former Member of the Competition Commission, and Chairman of the Dutch Electricity Market Surveillance Committee. He was President of the European Economic Association in 1996, was awarded the Frisch Medal of the Econometric Society, the Harry Johnson Prize of the Canadian Economic Association in 1993, and the recipient of the IAEE 2002 Outstanding Contributions to the Profession of Energy Economics Award. Dr. Newbery has managed a series of research projects on utility privatization and regulation, electricity restructuring, road congestion and road pricing, tax reform in Hungary and Greece. He has published over 100 academic journal articles. His two most recent books are A European Market for Electricity?, 1999, and Privatization, Restructuring and Regulation of Network Utilities, MIT Press, 2000. He was the Guest Editor of the special issue of The Energy Journal on European electricity liberalization in 2005. Dr. Newbery was educated at Trinity College, Cambridge with BA Degrees in Mathematics and Economics. He has a PhD in Economics and an ScD Degree from the University of Cambridge. He is Fellow of the Econometric Society and of the British Academy. Richard P. O’NEILL is the Chief Economic Advisor at the Federal Energy Regulatory Commission (FERC). Previously, he was the Chief Economist and Director of the Office of Economic Policy, the Director of the Commission’s Office of Pipeline and Producer Regulation at FERC, and prior to that, he worked at the Energy Information Administration. His work has focused on open access, restructuring, competition, performance-based incentive regulation, and market design. Prior to that he was on the Computer Science and Business Faculty of Louisiana State University and the University of Maryland. He has worked with several countries, states, the World Bank, energy companies and computer companies in the development of mathematical software, energy modeling, forecasting, regulation, privatization, restructuring and market design. His published work has appeared in academic and professional journals and books in the areas of applied mathematics, optimization, operations research, management science, computer science, energy, electrical engineering, economics, and law. He has a BS in Chemical Engineering, an MBA and a Doctorate in Operations Research from the University of Maryland. Yannick PEREZ is Associate Professor of Economics at the University of Paris-Sud 11 since 2003 where he is engaged in research in the field of Electricity Market Reforms, Industrial Organisation and Institutional Economics within the Groupe Réseaux Jean-Monnet at the ADIS laboratory. Prior to joining the faculty, he studied at University Paris Panthéon-Sorbonne. Dr. Perez’s main research interests include study of restructured electricity markets, benchmarking studies including study of company methodologies and strategies. He has published widely on these topics. Dr. Perez has Degrees in Economics from the University of Paris, Sorbonne and Ecole Normale Supérieure de Fontenay St-Cloud, and holds a PhD from the University Paris, Sorbonne. Wolfgang PFAFFENBERGER is Professor of Economics at the International University Bremen and Director of Bremer Energy Institute, an interdisciplinary research center focusing on energy system analysis and energy markets.

xiv

Contributors

He is especially interested in the economics of competition with reference to network industries and the electricity and natural gas industries in particular. He has published a textbook on electricity economics (in German) and co-authored a textbook on energy economics (in German) and written numerous studies and articles on macroeconomic implications of energy and environmental policy, on various aspects of market opening in the electricity supply industry (ESI) and on the special problems of the ESI in transformation countries. Dr. Pfaffenberger holds a Diploma in Economics from Freie Universität Berlin where he also got his PhD. He is Honorary Doctor of the Faculty of Economics of the State University of Novosibirsk in Russia. Ricardo RAINERI is Professor of Economics at the Engineering School of the Pontificia Universidad Católica de Chile, where he is engaged in research on industrial organization and regulation of the Energy Industry. Dr. Raineri’s main areas of interest include economics of competition, regulation, markets’ restructuring, and pricing. He has widely published in a variety of energy and related academic journals and authored two books. Also, he has been consultant for private companies and for the government, member of arbitrage commissions, and has been expert witness at the Chilean Congress. Dr. Raineri has Degrees in Business Engineering and Master in Economics from Pontificia Universidad Católica de Chile. He obtained a MA and a PhD in Economics from the University of Minnesota, under the supervision of Prof. Edward C. Prescott, 2004 Nobel Prize Laureate. Fereidoon SIOSHANSI is President of Menlo Energy Economics, a consulting firm specializing on the economics of the electric power sector. His professional experience includes working at Southern California Edison Company, the Electric Power Research Institute, National Economic Research Associates and Henwood Energy Services, Inc., now Global Energy Decisions. Dr. Sioshansi’s main research and consulting interests include electric power restructuring and market liberalization, global climate change, environmental and energy policy, sustainability, demand and price forecasting, integrated resource planning, energy efficiency, and renewable energy technology. He has authored numerous reports, books, book chapters and articles. He is the Editor and Publisher of EEnergy Informer, a monthly newsletter on the electric power industry. He is a frequent contributor to The Electricity Journal and Energy Policy and is on the editorial board of Utilities Policy. He has Degrees in Engineering and Economics, including an MS and PhD in Economics from Purdue University. James L. SWEENEY is Professor of Management Science and Engineering at Stanford University; Senior Fellow of the Stanford Institute for Economic Policy Research; Senior Fellow of the Hoover Institution on War, Revolution and Peace; and Senior Fellow of the Stanford Institute for International Studies. His professional activities focus on economic policy and analysis, particularly in energy, natural resources, and the environment. At Stanford he has served as Department Chair, Director of the Energy Modeling Forum, Chairman of the Institute for Energy Studies, and Director of the Center for Economic Policy Research (now the Stanford Institute for Economic Policy Research). In the early 1970s he was Director of the Office of Energy Systems Modeling and Forecasting of the US Federal Energy Administration. He was a Founding Member of the International Association for Energy Economics and Co-editor of the Journal Resource and Energy Economics. He is Fellow of the

Contributors

xv

California Council on Science and Technology and a Member of Governor Schwarzenegger’s Council of Economic Advisors. He holds a BS Degree from MIT in Electrical Engineering and a PhD from Stanford University in Engineering Economic Systems. Michael TREBILCOCK is the Director of the Law and Economics Program at the University of Toronto Faculty of Law. His research and teaching interests over the past 36 years relate to the fields of competition policy, contract theory, international trade regulation, and economic and social regulation, all areas in which he has published widely. His numerous academic honors include the 1998 Molson Prize for contributions to the social sciences in Canada. Over the past 20 years, he has consulted widely to government and the private sector on matters of competition policy and economic and social regulation. Prof. Trebilcock was Research Director to the Electricity Market Design Committee during 1998, appointed by the Government of Ontario to design the detailed rules for the introduction of wholesale and retail competition in the ESI in Ontario in 2002. He holds an LLB Degree from the University of Canterbury, New Zealand, and an LLM Degree from the University of Adelaide, South Australia. Taff TSCHAMLER is Senior Principal at KEMA and Director of the firm’s North American retail energy practice. His primary areas of expertise include business planning, market analysis, policy assessment and financial and risk analysis of retail energy markets. He has advised numerous North American utilities, competitive energy suppliers, private equity firms and large energy buyers on market and regulatory strategy. He has published in leading industry periodicals such as The Electricity Journal and Power Economics. Prior to joining KEMA he was with Decision Analysis Corporation and ICF Consulting. Mr. Tschamler holds a Masters Degree in Public Policy from the College of William and Mary and a BA in Economics with distinction from the University of Maine. Masayuki YAJIMA is Executive Research Economist at Central Research Institute of Electric Power Industry (CRIEPI) in Japan where he is engaged in research on the electric power industry. Dr. Yajima’s research is concentrated in areas of regulatory reform, management strategy of utility companies to cope with competition as well as energy security policy. Dr. Yajima has published many books on these topics, including Deregulatory Reforms of the Electricity Supply Industry, Liberalization of Electricity Markets, Electric Restructuring, Big-Bang in Electricity Markets in the World, Energy Security, and Reconsideration of Electric Restructuring. Dr. Yajima holds a Masters Degree in Public Administration as well as a PhD in Public Administration from International Christian University in Tokyo. Jay ZARNIKAU is President of Frontier Associates engaged in retail market strategies, utility pricing, demand forecasting, and energy policy. He formerly served as a Vice President at Planergy, and prior to that, he was with The University of Texas at Austin Center for Energy Studies, and worked at the Public Utility Commission of Texas. His publications include articles in The Energy Journal, Resource and Energy Economics, Energy Economics, IEEE Transactions on Power Systems, Energy Policy, Energy, and The Electricity Journal. His past papers have estimated demand response in the ERCOT market, market concentration in ERCOT, and examined cogeneration policies in Texas. Dr. Zarnikau has a PhD Degree in Economics from The University of Texas at Austin, where he also teaches Applied Statistics as a part-time visiting professor.

This page intentionally left blank

Foreword: The Market versus Regulation STEPHEN LITTLECHILD University of Birmingham, UK; Judge Business School, University of Cambridge, Cambridge, UK

From its beginnings in Chile and the UK in the 1980s, electricity reform has spread almost throughout the world, especially in the more developed countries. The objective has generally been to increase efficiency, reduce costs, and improve quality of service. The reforms have varied in extent and scope and detail but in general they have sought to achieve this objective by relying less on public enterprise and regulated monopoly, and more on market mechanisms such as private ownership and competition. With up to two decades of experience now in hand, discussion has increasingly gone beyond theory to look at practice. Many books and papers have analyzed experience in individual countries, especially the UK and USA. There have been some comparative studies within wider geographical areas, such as Europe. But the present volume is perhaps the first, certainly the most extensive and up to date, systematic comparison of experience on a worldwide basis with the aim of identifying what works and what does not and why. The editors have invited me to focus this Foreword on the role of the market versus regulation. Each chapter in the book contains a wealth of relevant information, analysis, and stimulating interpretation, but it would be impossible to discuss and evaluate all the chapters here. Paul Joskow’s Introduction in any case does much of this job. It is a comprehensive, authoritative, indeed magisterial survey of experience worldwide. It reflects his incomparable research record and practical involvement in reform, and exhibits considered judgment. I cannot hope to better it, and I agree with it almost entirely, although we may differ on a few aspects of policy. I shall begin by briefly reiterating the aims, criteria, textbook model, and main lessons to be learned. These lessons seem to me deserving of general acceptance. Then I want to discuss four major topics – competition in the wholesale and retail markets, the nature of network regulation, and the regulation of transmission networks – where we observe not only variation in practice but also differences of view in the literature. Here I want to argue that there is a greater role for the market vis-à-vis regulation than is often accepted.1 The Aims and Assessment of Reform Proponents of electricity reform have had many and diverse aims, not always mutually consistent. The Introduction suggests that “the over-riding reform goal has been to create new 1

For simplicity of exposition, this Foreword does not include references to the literature. Relevant references can be found in my Beesley lecture which covers some of the same topics in more depth “Beyond Regulation”, Beesley Lecture Series XV, October 4, 2005 (revised version October 12, 2005) at http: //www. electricitypolicy.org.uk/pubs/misc/Beesley.pdf, forthcoming in Colin Robinson (ed.) (2007), Government and Utility Regulation (provisional title), Cheltenham: Edward Elgar. I am grateful to Paul Joskow for helpful comments on this Foreword, to Jim Sweeney for clarifying some points on California, and more generally to many other colleagues acknowledged in the Beesley lecture.

xvii

xviii

Electricity Market Reform

governance arrangements that provide long-term benefits to consumers”. These benefits are to be realized by creating competitive wholesale and retail markets to improve efficiency and responsiveness to customer preferences, by incentive regulation of privatized transmission and distribution networks to improve their efficiency and facilitate competition across them – and, I would add, by reducing the role of government and political influence generally. This fairly summarizes the aims that many of us had in the UK. I believe it also reflects the main aims in those other countries that embarked on thorough-going reform of the electricity sector. (In general this aim was not limited to the electricity sector, of course.) In assessing performance, it is necessary to adopt a “comparative governance” approach. Observed performance should be compared against clearly defined alternative institutional arrangements, recognizing that “ideal” textbook performance is never achievable in reality. In this respect, the “new institutional economics” is an important advance on the prevalent economic analysis of, say, 30 years ago. But economists have sometimes been reluctant to abandon the ideal theoretical benchmark (what Demsetz called the “nirvana fallacy”). Even now, public choice considerations do not always inform the analysis of regulatory alternatives. The “Textbook Model” for Restructuring and Competition What I have elsewhere called the “standard model” for electricity reform is well spelled out in the 10 components of the “textbook architecture” for restructuring and competition set out in the Introduction. In summary, they are: ●

●

● ●

●

●

●

● ●

●

Privatization to enhance performance and reduce the ability of the state to use these enterprises to pursue costly political agendas. Vertical separation of competitive and regulated monopoly sectors to facilitate competition and regulation. Horizontal restructuring to create an adequate number of competing generators and suppliers. Designation of an independent system operator to maintain network stability and facilitate competition. Creation of voluntary energy and ancillary services markets and trading arrangements, including contract markets and real-time balancing of the system. Application of regulatory rules to promote access to the transmission network and incentivize efficient location and interconnection of new generation facilities. Unbundling of retail tariffs and rules to enable access to the distribution networks in order to promote competition at the retail level. Specification of arrangements for supplying customers until retail competition is in place. Creation of independent regulatory agencies with adequate information, staff and powers, and duties to implement incentive regulation and promote competition. Provision of transition mechanisms that anticipate and respond to problems and support the transition rather than hinder it.

I am tempted to add a final component: Do nothing more. At least, the need to avoid excessive government and regulatory involvement is one of the lessons to be learned. The Importance of Following the Textbook Model Where the “textbook model” has been largely followed it has been broadly successful; for example, in the UK, Argentina, the Nordic countries, Victoria, and Texas. Where it has not been followed, there have been problems. Departures from the textbook model include sins of omission and sins of commission, and in some cases both.

Foreword: The Market versus Regulation

xix

Examples of the sins of omission include Belgium, Italy, and especially France, where relatively little has been done to restructure to create competitive markets (and in France’s case to privatize). Opening up to competitive markets seems to have been very slow in Japan. In Germany and New Zealand there were initial failures to recognize the need for a sector regulator. Examples of the sins of commission include Chile (excessive restrictions on generation and competition) and California (inappropriate restrictions on contracting and retail pricing and undue government involvement). Examples of both types of sins include most other North American jurisdictions, where there has often been insufficient restructuring coupled with excessive retail (and in some cases wholesale) price controls. In Ontario these problems have been compounded by undue government involvement. Similarly, continued state ownership in some states in Australia has been coupled with excessive retail price controls. In many of these cases, competition has been less effective, and prices to customers have been correspondingly higher, than would have been the case had the textbook model been adopted in full. In other cases, by contrast, prices have been artificially held below market levels, which has been the cause of different problems. The inability or unwillingness of governments to secure and defend market prices that cover reasonable costs has often precluded the full application of the textbook model. This has been the case in many developing countries, particularly in Asia.

California The problems in California have been much analyzed, not least in this volume. But since they are cited around the world as a reason for not engaging in electricity reform, it is worth a few words here to explain why that is the wrong lesson to draw, and why experience in California should not be a deterrent to electricity reform elsewhere. Commentators in this volume are right to explain that the California problems were not primarily a failure of the wholesale market. Generators wanted to invest but were slowed down by delays in regulatory approvals (which did not reflect environmental restrictions) and by regulatory uncertainty until restructuring policy was clarified. There may have been some exploitation of generator market power (to an extent that is subject to debate). But this was exacerbated, if not caused, by the regulatory framework. The main mistake was to prohibit or discourage the incumbent utilities (as retail suppliers to the majority of customers) from entering long-term contracts with the generators. This almost invited increases in spot market prices. Another problem was the obligation on the utilities – once they had covered their stranded costs and the price caps had expired – to pass directly through to customers the prices obtaining in the wholesale market. This happened in San Diego, and when prices rose sharply in the wholesale market the extent of customer protests in San Diego led directly to the intervention of the Governor of California. A third problem was the inflexible retail price caps that led to bankruptcy or near-bankruptcy for the other two utilities when wholesale prices rose. These caps did not incorporate the costs of entering long-term contracts to hedge against such price rises because the utilities had been discouraged from entering these. The refusal to raise the caps exacerbated the financial problems of the main utilities and reduced any role that reductions in demand could have played in ameliorating the electricity crisis. The main problem, in short, was one of inappropriate regulation, and was not attributable to privatization or competitive markets per se. In partial defence of the regulatory commission, it

xx

Electricity Market Reform

has to be admitted that the main components of the policy in question (no contracts and wholesale spot price pass through) were ones that had been advocated by leading US economists and advisers at the time. It is now generally accepted that neither of these components is appropriate, and that suitable contracting is to be encouraged. Of course, the situation in California was much exacerbated by political involvement. The then-Governor refused to allow any retail price increases in the rest of the state and declared an emergency situation. The California Legislature required that the Department of Water Resources purchase electricity on behalf of customers, which it did via large quantities of very long-term contracts at what turned out to be excessive prices. The Legislature also suspended retail competition (direct access) in order that the costs of these contracts could be passed on to retail customers. The California Public Utility Commission voted to implement this (with a strongly dissenting minority). I return to some of the retail competition issues later. But the point to emphasize here is that the proper lesson of California is to avoid over-regulation, not to avoid electricity reform. Competition in the Wholesale Generation Market International experience supports the argument for dealing with potential market power ex ante rather than ex post, and for doing so structurally rather than by restrictions on conduct. At the time of electricity privatization in the UK there was an awareness of the need for several generators from the outset, but this was outweighed by the desire to privatize the nuclear stations as well. After the nuclear stations were withdrawn, a concern to meet the privatization timetable precluded remedying this mistake. It was not straightforward or costless to remedy the situation later. Argentina, Australia, and some other countries learned the lesson and initiated a more extensive and beneficial structural separation. Much of continental Europe has not yet done so. Stronger transmission links between countries are now seen as the solution. It remains to be seen whether this will suffice to alleviate market power and political influence within each country. Most US jurisdictions have instituted market monitoring and mitigation protocols. Much thought and expertise has gone into these. But sometimes they seem rather prescriptive and severe. The Introduction reflects a concern that they may have constrained prices from rising to competitive levels, especially for peaking plant, thereby creating adverse investment incentives in the longer term. Some market monitoring protocols may be a regulatory over-reaction to early price spikes. Arguments by some economists that observed prices reflected market power because they exceeded short-run marginal cost were not helpful here. It turns out that the prices in question were often barely above the cost of staying in the market, and generally below the longrun cost of entering the market. This illustrates the disadvantages of using an inappropriate theoretical benchmark. Is it possible that concerns about market power may have been exacerbated by the wholesale market design? A characteristic of an electricity pool is that the price becomes a kind of public good, imposed on all parties uniformly without their explicit agreement (other than their acceptance of the process). This may have some advantages but the price can also become a matter of public concern. Attention and complaints naturally focus on it. An advantage of bilateral trading (as in the UK) is that there is no single imposed price and all trades are voluntarily agreed between the parties. It is encouraging that many pools are

Foreword: The Market versus Regulation

xxi

now allowing bilateral trading outside the pool, and it will be interesting to see how both forms of market develop. Adequacy of Generation Investment A different concern is now emerging in some countries, under the heading of security of supply or resource adequacy. Can competitive markets stimulate adequate investment in new generation? Some policy-makers in particular fear that they may be held responsible for any deficit. Whether these concerns are justified is another matter. The market has already brought forth extensive investment in many different countries. There is no reason to expect that it will not be equally forthcoming in future, provided that it is not discouraged by inappropriate regulatory or government policies. As the Introduction explains, such potentially discouraging policies include macroeconomic and political instability; the uncertainty, complexity, and discouraging effects of investment by state-owned entities (including possible or threatened investment); the similar effects of some programs to stimulate uneconomic renewable sources; the costs and uncertainties of continuous market redesign; and actions by regulators or independent standard operators (ISOs) that unduly limit or depress market prices. Where such policies are discouraging generation, the remedy is surely to cease or moderate them rather than to conclude that the market does not work. Many chapters in this volume document significant investment in new generation. Are some of the authors unduly optimistic about the future? It is true that the UK has benefited from the ability to call on mothballed generation plant (new as well as older plant). It is surely an achievement of privatization and competition to have induced generating companies to consider mothballing as an economic option where it was previously ruled out. There seems no reason why other countries should not follow suit where it is economic to do so. Importantly, however, mothballing is not at the expense of new plant: there has also been and still is significant declared intention to construct new plant in the UK and no doubt elsewhere. There may be less interest in new generation in some countries if prices do not justify it at the moment. But low prices are not inherent in a competitive market, unless the market rules and monitoring guidelines so dictate. There is a danger that what emerges from the market are the prices that the regulator or market monitor deem appropriate, rather than what a competitive market would require in order to meet demand. Again, the solution may be to reconsider the nature and scope of the market rules and the monitoring guidelines. As noted earlier, bilateral trading also offers advantages here. Some countries or jurisdictions where a wholesale market already exists are implementing or considering a greater role for regulation; for example, by imposing forward contracting obligations or administratively determined capacity payments. This is disappointing: I regard it as a step forward to have abolished such a capacity mechanism in the UK. Regard must be had to the way such arrangements would work in practice rather than in theory. Market participants are quite capable of manipulating any such regulation to suit their own ends. It is also necessary to be realistic about regulation. Are regulators really capable of forecasting future demand and determining what levels of contracting obligations or capacity payments are needed to meet this demand most economically? And is this what in practice will influence their decisions? Introducing more regulation involves creating a greater role for political considerations, which are invariably influenced by a variety of factors, often of

xxii

Electricity Market Reform

a short-term nature. Past political concerns have included the levels of costs and prices and the extent of competition; today the concerns include resource adequacy and the environment; what will be the concerns tomorrow?

Competition in the Retail Supply Market We now understand more clearly the problems that emerge for retail competition if an inadequate regulatory framework is put in place. Defects include no or inadequate unbundling and ring-fencing of distribution networks and retail supply, inadequate provision for equal access to networks on non-discriminatory terms, cross-subsidization of one service or supplier by another, unclear allocation of costs and setting of price controls, inadequate provision for revenues to cover default services, use of default services or price controls to achieve objectives other than a robust retail market, and so on. A failure to recognize these problems has often proved critical in practice, sometimes fatal. Some might suggest that a lack of retail competition does not matter. They see costs but not many benefits. The Introduction notes that the potential benefits of retail competition include lower prices (from lower costs of energy purchasing and retailing, admittedly perhaps offset by increases in some other retailing costs), and new value added services such as risk management, demand management, and energy services. No one disputes that these benefits are likely to outweigh the costs for large and medium industrial consumers. Failure to provide for effective retail competition in this sector of the market is generally agreed to be a serious error, and policy is coming to reflect this internationally. However, some commentators are not convinced that this is the case for residential and small commercial customers. Their concern is that retail competition for such customers introduces additional costs and higher profit margins, and therefore is only feasible if regulated or default prices are increased to more than cover these additional costs and margins. These commentators prefer, and consider that customers would prefer, a regulated alternative to a competitive market. I take a different view. Although regulation may seem to offer lower costs and prices in the short term, I believe that the market will offer better value in the longer term when one considers how regulation will actually operate. Let us first identify those markets where residential retail competition has so far developed relatively well, note some characteristics of those markets, then examine what forms of regulation are realistically available.

International Experience of Retail Competition in Residential Electricity Markets Table 1 lists the main residential markets that are now open. Some of them exhibit significant switching away from the incumbent supplier, but others do not. Of those residential markets that opened about 6 years ago (in the period 1998–2000) the proportions of residential customers with non-incumbent suppliers are now 43% in UK; 29% in Sweden, and 24% in Norway but only about 11% in Finland and 5% in Germany; 26% in New Zealand but seldom over 7% in North America. In a few US states some high proportions were observed initially, but in only a few territories in those states. Of those residential markets that opened just over 3 years ago (in January 2002) the proportions are already 24% in Texas and 33% in Victoria though only 11% in New South Wales. In Ontario (which opened in May 2002) 20% of customers had signed with another supplier

xxiii

Foreword: The Market versus Regulation Table 1. Residential customer switching in international electricity markets.2 (proportion (%) of residential customers served by non-incumbent supplier). Market Markets opened 1998–2000 UK Sweden Norway Finland Germany New Zealand Alberta California Maine MPS BHE & CMP Maryland Potomac Electric Other utilities (3) Massachusetts New Jersey New York Ohio First Energy (3) Cincinnati Other utilities (4) Pennsylvania Duquesne Light PECO Energy Other utilities (4) Markets opened January 2002 Texas Ontario (open May 2002) New South Wales Victoria

After approximately 3 years

After 5–6 years

34 18 15 5 4 18?

43 29 24 11 5 26

2 2

7 1

36 0

7 0

15 0 3 0 4

6 0 3 0 6

40 2 0

45 3 0

35 18 1–7

23 2 1

19 (now 24) 23 (in September 2002, then 0) 9 (now 11) 24 (now 33)

by the day the market opened, but within a few months a price cut and subsidy by the state government led to the disappearance of the market.3 With the exception of most North American markets, the proportions of customers switching are growing steadily over time; in North America (apart from Texas) the proportions are generally static or declining. The high switching markets – notably the UK, Norway and Sweden, New Zealand, Victoria, and Texas – exhibit many other forms of competition as well. They are generally characterized by considerable entry and exit of suppliers; by growth and decline of individual suppliers; by 2

My Beesley lecture gives sources and further discussion. Whether any competitive contracts remained valid is unclear. Competitive offers (for 3- and 5-year fixed-price contracts) are now beginning to reappear in Ontario. 3

xxiv

Electricity Market Reform

mergers and takeovers; by a variety of marketing techniques; by active competition on price; and by an increasing variety of non-price services and product variations.4 Of particular interest is the increasing range of contractual terms. In addition to the traditional standard tariff that is variable at the utility’s (or regulator’s) discretion, the Nordic markets offer a wide range of products including fixed-price contracts varying from 3 months up to 5 years, and spot price-related contracts including a variety of optional hedges. Up to 40% of residential customers have chosen such contracts. They have exhibited a wide range of different preferences, which have also responded to changing market conditions and evolved over time. These retail electricity markets seem increasingly indistinguishable from other competitive markets such as banking, insurance and mortgages, other fuels including petrol (gasoline) and heating oil, telecommunications, food and housing, and indeed many consumer goods and services generally. These markets too involve costs to operate, some of which could no doubt be reduced if there were a single regulated supplier. Consumers there too may be “sticky”, with some reluctant to change from their traditional supplier. From time to time there may be concerns about some aspects of these markets, and a variety of restrictions may be applied to suppliers in them. But in general economists do not consider that it would be better to replace competition in these markets by a regulated outcome. Have we somehow discovered the one product in the whole of the consumer market for which regulation is better than competition?

Alternative Mechanisms or Regulating Residential Markets Consider now what form the regulation of residential electricity markets might take. The policy might be “a regime where the distribution company procures power competitively and resells it at cost”. But this sounds disconcertingly like the ideal theoretical benchmark. If nationalized industries or regulated public utilities could and would do this there would be no need for privatization and competition in the first place. Socialism and communism would work better than markets. What are the practical regulatory alternatives to retail competition? They include the following: ●

●

●

●

4

The regulator specifying ex ante the quantities and timing and prices of energy purchases that are to be made, and the terms on which this energy should be sold to customers, or approving a detailed process for doing this. The regulator approving or disapproving the above items ex post, and consequently allowing or disallowing the costs and revenues involved. The regulator using benchmarks based on purchasing by other comparable suppliers to use either ex ante or ex post in the above schemes. The regulator setting price caps for specified periods of time based on assumptions about the costs of hedging variations in energy prices.

In the UK, these variations include bundled offers notably dual fuel, credits in the form of airmiles, loyalty points with specific retailers or shopping cards (Nectar), contributions to charities and deserving customer groups, green tariffs, energy efficiency packages, insurance cover, discounts for self-reading meters, the Staywarm scheme offering unmetered electricity for a fixed monthly fee, discounts for various prepayment meter schemes, discounts for a range of home services and financial products, tariffs with no standing charges, single billing for up to six utility and other services, and so on.

Foreword: The Market versus Regulation

●

xxv

The regulator specifying or approving ex ante the basis on which contracts for supply are to be put out to tender, and the prices at which this supply is to be priced to customers over time.

All these possibilities are “workable”, and all have probably been tried at one time or another. I myself have been involved in several of them. Some have greater merit than others. Fixedprice caps worked acceptably in the UK for a transitional period of 4 years. Price cap adjustment mechanisms have so far not been inconsistent with competition in Texas. Competitive tendering seems to have secured very good prices for customers in New Jersey and Maine, at least in the short term. But all these alternatives have disadvantages too. For example, it is easier to envisage a comparator for an incumbent supplier in Holland than in France or Italy. And even with several suppliers in areas of similar size the variations between suppliers (e.g. in customer mix, consumption patterns, weather conditions and other factors) should not be underestimated.

The Decision-Making Process Let me focus here on two aspects of the institutional comparison between regulation and the competitive market. The first aspect is the decision-making process, whether by the regulator or by market participants generally, and the information available for this purpose. I refer here not to information about what the utility companies are doing but about what the future holds for wholesale prices. Whoever is purchasing electricity, or prescribing or approving its purchase ex ante or ex post, has to decide when and what to buy. Is it better to buy ahead or on the day? In the former case, is it better to buy a week ahead, or 3 months or a year or 3 years or 15 years ahead? What should the portfolio of purchased contracts look like? Should they be fixed-price or indexed contracts, and if indexed to what input price or other parameters? There are even more far-reaching decisions to be made. Is it better to integrate vertically into generation rather than buy on the wholesale market? If so, to what extent is it prudent to integrate and what kinds of generation plant is it best to buy? There is a comparable set of decisions on the selling side, with analogous required information. Should the electricity be sold at a price that varies according to wholesale market conditions? Or should there be fixed prices and if so when should they be set and for what periods? A tender for 2 years might imply fixing the price for 2 years – but is 2 years the right period, and does this depend on views about the future level of prices? British Gas has just offered an optional fixed price for 5 years, but in Norway suppliers are now changing tariffs as frequently as monthly to reflect changing wholesale market conditions. If a single product is to be offered, which is the right one? If, instead, customers should be given a choice, how to discover what customers want and to predict what they will choose, and how to adapt to their perceptions and preferences that evolve over time? These are decisions that have to be taken in any market, and they are increasingly important in the electricity market. Economists normally argue that it is better to allow many players to take such decisions in a competitive market than to give a regulator monopoly power over such decisions. It is not that private suppliers invariably get the decisions right or that regulators are particularly unintelligent or unwise. Rather, competition is a discovery process that itself tends to identify and encourage those individuals and organizations that prove better at purchasing inputs and at understanding and providing what customers want. Competition also tends to eliminate those decision-makers that prove less able in these respects. There is no

xxvi

Electricity Market Reform

comparably effective discovery process for regulation. Do not the principles of competitive markets apply to electricity too? Retail Competition and Public Choice The other aspect of the institutional comparison between competition and regulation concerns how regulation actually operates. Here, both statutory and public choice considerations must be brought to bear. Is a regulator in practice able to purchase power competitively and sell it at cost? The duties of a regulatory body will typically require it to have regard to a wide range of considerations, directly or indirectly including some reference to the public interest, and generally referring to the long term as well as the short-term interests of customers. These duties would apply to the regulator in assessing what kinds of contracts the utility should purchase, and at what prices, and on what basis the power should be sold on to customers. In addition, politicians, the media, interest groups and not least the government will from time to time draw regulators’ attention to the importance of this or that consideration. Yesterday the main consideration may have been coal or promoting competition or reducing price, today it may be renewables or security of supply, tomorrow it might be demand management or nuclear. Similar considerations apply to decisions about the terms on which a regulatory body should authorize electricity to be sold to customers. The same parties will ask: Is it really appropriate to increase consumer prices just at this moment? And is it really appropriate to offer products that distinguish between this and that type of customer? These are the kinds of broader- and longer-term considerations that seem to me crucial in assessing the case for retail competition. If the aim is to facilitate certain kinds of political or non-market objectives, then regulation may have advantages over retail competition. But decisions of regulators and governments as an alternative to retail competition can be very costly to customers: witness the tens of billions of dollars associated with the coal contracts in the UK, the renewable and nuclear purchasing policies that first prompted reform in California, and the more recent insistence on expensive long-term contracts that nearly ended reform there. If the aim of electricity reform is to create governance arrangements that provide long-term benefits to consumers, then retail competition down to the residential level should be an integral component of the textbook model. The Nature of Network Regulation Incentive regulation has been extensively developed and applied in the UK and many other jurisdictions. It has led to increases in efficiency, reductions in prices, and improvements rather than reductions in reliability. In the UK, at least, there has been more capital investment than in the previous regime, not less. Fears that the mechanism is inconsistent with adequate security of supply have proved groundless. In general, consumers have benefited and so have investors. In contrast, it seems that consumers in Germany and New Zealand have suffered from the absence of such regulation. US regulatory commissions have not shown much interest in effective incentive regulation. They do not seem to have gone beyond accepting voluntary price freezes. It has been said that US regulators have no power to impose an incentive price cap. It may be that the US regulatory framework, with the obligation on the regulator to prove in course of litigation that a particular expense is unnecessary or a particular investment is not required, is not conducive to the approach.

Foreword: The Market versus Regulation

xxvii

This makes it particularly important to look at alternative mechanisms for achieving mutually advantageous outcomes. In reality, negotiated settlements between utilities and interested parties, subsequently approved by regulators, have been an important feature of US regulation. I understand that Paul Joskow noted this in his thesis many years ago. But economists have almost completely neglected this practice. Recent research shows that the informal or negotiated settlement process differs fundamentally from the formal litigation process and has significantly different outcomes. For example, Federal Energy Regulatory Commission (FERC) has accepted settlements for the majority of gas pipeline cases. The typical outcomes have been rate moratoria that FERC itself could not impose. In Florida, the Office of Public Counsel (the consumer advocate) has gone further. It has negotiated with the utilities over three quarters of the electricity rate reductions that have been achieved over the last quarter century. This amounted to nearly $4 billions for consumers. For their part, the utilities gained more flexibility in accounting procedures, and also more attractive fixed-term revenue-sharing arrangements that provided greater incentives to efficiency than conventional rate of return regulation. The possibility of negotiated settlements instead of regulatory-determined price controls has some appeal even in jurisdictions where there is no barrier to incentive regulation via price caps. The advantages are at least two-fold. 1. First, the outcomes would reflect the preferences of the parties themselves, rather than those of the regulator. This could be particularly important in issues such as investment for reliability of supply. Certainly the Electricity Consumers’ Committee in the North of Scotland took a different view from the UK regulator in 1995, siding with the utility’s argument for more investment despite the higher price. On appeal, the Monopolies and Mergers Commission endorsed the view of the utility and the consumer committee on this issue. 2. The second advantage is that negotiated settlement potentially introduces greater variety of outcomes; for example, different kinds and durations of price caps. The regulator is effectively constrained to a uniform approach at any time. Negotiated settlements therefore offer more scope for learning from experience. The details remain to be worked out in each jurisdiction. There might need to be a regulatory backup, or at least an appropriate regulatory context. However, the principle of negotiated settlements seems worth further exploration, including in those jurisdictions that have already been keen to implement incentive regulation.

The Regulation of Transmission Networks The Introduction describes the relative lack of transmission investment in several countries, especially the USA and Europe; notes the adverse effects this has on congestion, competition, reliability, and cost; contrasts the comprehensive arrangements in the UK and Argentina with the less considered approaches in many other countries; suggests that relying primarily on market-based merchant transmission is likely to lead to inefficient transmission investment; and commends the progress being made in certain other countries. I am not well able to assess all these issues, but most of the above diagnosis is consistent with my understanding of the situation in many countries. However, I have been especially concerned about the analysis of “market versus regulation” in the context of transmission.

xxviii

Electricity Market Reform

The case for merchant transmission seems to have depended on the theoretical argument that, under certain conditions, it will lead to optimal investment. The case against merchant transmission is that these conditions are unlikely to be met. In simple terms, the main concern is that, because of market power and other market failures, merchant transmission is likely to be “too little, too late”, if indeed it happens at all. However, as noted earlier, the relevant benchmark is not some theoretical optimum. It is necessary to consider whether these theoretical market failures are serious in practice. It is also necessary to compare each proposed solution against the institutional arrangements that would be adopted instead. Suppose the alternative to merchant transmission is the conventional decision-making process by a transmission company and/or a regulator, with costs assigned to other parties (transmission users) regardless of outcome. There is no obvious reason why a regulated entity will be better able to predict demand. And public choice theory, backed up by considerable evidence, suggests that there will be commercial and political pressures to build excessive transmission capacity, or “too much, too soon”. In other words, there is a possibility of regulatory failure. It is therefore necessary to compare the alternatives as they would work in practice. Which is likely to exhibit the more serious failure?

Merchant and Regulated Interconnectors in Australia Experience in Australia has been an eye-opener. Both merchant and regulated transmission lines have been built. All are interconnectors between electricity regions rather than expansions within a single system. However, they clearly illustrate some of the important factors at work. Two merchant lines have been built. They seem to have underestimated the speed and extent to which new generation in the high price regions would reduce the price differentials between the regions. This means they overestimated the economic benefits and profitability of the lines. Their market power has in practice been negligible. Far from being “too little, too late” they both appear to have been “too much, too soon”. Their investors have had to foot the bill for the misjudgments of the value of the interconnectors. They have learned their lesson and settled for a regulated income. One regulated line has been built broadly in parallel to the first merchant line. Arguably it misjudged the demand even more severely: it was about five times the size of the merchant line and correspondingly much more uneconomic. Transmission users rather than investors are having to foot this bill. Separately, another transmission company and the regulatory body would have built another regulated line, duplicating the second merchant line and being wholly redundant, if the courts had not stopped them from doing so. What has been or would be the experience of transmission expansions within an electricity system, sometimes called “intensive network upgrades”, deserves further research. But in the case of interconnectors between electricity regions in Australia, there seems no doubt that regulatory failure has been more serious than market failure.

The Public Contest Method in Argentina An alternative to both regulated and merchant transmission is the Public Contest method used in Argentina. Transmission expansions have to be proposed, approved, and financed by users themselves. Construction is then put out to tender. The users typically comprise generators, distribution companies, and large industrial consumers. This arrangement was carefully designed to avoid the inefficient over-expansion that characterized the pre-privatization

Foreword: The Market versus Regulation

xxix

era in Argentina. It was envisaged, correctly, that if users who benefited from an expansion had to pay for it as well, they would give more careful consideration to whether it really was worthwhile. Although the approach has received some unjustified criticism, it has actually worked rather well. The main concern was that users initially turned down a long-expected Fourth Line bringing power to Buenos Aires. But on closer examination the line turns out to have been uneconomic. Nowadays it is cheaper to transport gas to Buenos Aires and to generate power there. The Public Contest method has a good record. It enabled a variety of economic expansions to go ahead, halved the cost of building transmission lines, greatly increased the productivity of the transmission system, and resisted the political demands for uneconomic expansions in outlying regions.

Analyzing Institutional Arrangements for Transmission My argument is not that merchant transmission or the Public Contest method is necessarily the best approach in all circumstances. Nor do I propose no role for regulation. Rather, the economic analysis needs to be more consistent with the comparative institutional approach. The proposed rules about performance assessment (see Introduction) need to be applied in designing transmission (and other) arrangements ex ante as well as in assessing them ex post. Lessons also need to be learned from actual experiences. The examples I have cited suggest that the concerns about merchant transmission and user-determined expansions may not be as great in practice as feared in theory. And the alternative of regulated transmission has its own disadvantages that need to be taken into account. The chosen institutional mechanism must detect and prevent those transmission expansions that are excessive or uneconomic as well as discover and implement those that are needed and economic. There is indeed a great deal of work to be done on the creation of effective institutional arrangements to achieve this.

Conclusion To be successful, electricity reform must reflect a good understanding of what makes markets work well and what prevents this happening. Experience and analysis of many electricity markets over the last two decades is leading to broad agreement on the basic prerequisites. There is still scope for debate on the relative roles of regulation and the market. I have argued that regulation can usefully be reduced and the role of the market increased, not only at the wholesale and retail levels but also with respect to monopoly networks. The contributors to this volume have made a substantial contribution to advancing our understanding of all these issues.

This page intentionally left blank

Introduction to Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies PAUL L. JOSKOW1 Department of Economics, Centre for Energy and Environmental Policy Research, Massachussetts Institute of Technology, Cambridge, USA

Introduction During the 1990s, many countries began to restructure their electric power sectors with the goal of improving sector performance.2 The restructuring programs have included privatization of state-owned enterprises, the separation (ownership or functional) of potentially competitive segments (generation and retail supply) from segments that have natural monopoly characteristics and are expected to continue to be subject to price and entry regulation (distribution and transmission), the creation of competitive wholesale and retail markets, and the application of performance-based or incentive regulatory mechanisms (PBR) to the remaining regulated segments to complement traditional cost-of-service regulation. We now have a significant amount of experience with the performance of these restructuring, regulatory reform and competition programs. Evaluating the performance record is useful in order to provide guidance to policymakers considering comprehensive electricity sector reforms where they have not yet occurred, to guide refinements in the reform programs aimed at improving performance where they have already been implemented and, importantly, to provide factual evidence to respond to questions about sector performance from policymakers who continue to be skeptical about the value of at least some of the reforms that have been implemented or are being proposed. This book contains studies of the electricity sector reform programs, their consequences and remaining policy issues for a large number of countries. The studies cover countries with diverse characteristics. Both developed (e.g. Britain, Central Europe, the Nordic countries, US, Canada, New Zealand) and a few developing countries are covered (e.g. Columbia, Argentina, Brazil). It covers countries or states/provinces that entered the reform process with primarily privately owned regulated vertically integrated utilities (e.g. US, Japan, Germany, Alberta) and countries or states/provinces whose electricity sectors were dominated by state-owned

1

I am grateful for research support from the MIT Center for Energy and Environmental Policy Research and the Cambridge-MIT Institute. Comments from Stephen Littlechild were very helpful. This paper draws heavily on Joskow (2000b, 2003, 2005a, 2006). 2 A few countries started comprehensive restructuring programs earlier (e.g. Chile and the UK).

1

2

Electricity Market Reform

monopolies (e.g. Britain, much of Central Europe, Argentina, the Nordic countries, Ontario).3 The nature of the reform programs has also varied quite widely, from comprehensive “textbook” restructuring, regulatory reform and competition initiatives (e.g. Britain, Argentina, Texas, New Zealand, Alberta) to those that have tried to introduce competition with minimal structural or regulatory reform (e.g. Germany, France, Japan) and examples of reform programs that fall between the two extremes (e.g. Chile, California, the states within Pennsylvania–New Jersey–Maryland regional transmission operator (PJM)). Since the reforms were implemented at different times and with different rates of change in different countries, the experience with meaningful reforms that we can draw upon varies from more than 10 years (e.g. Britain, Argentina, Chile) to much shorter time periods (e.g. Texas, Germany, Japan, Ontario, Brazil). In all of the countries covered, the initial reforms were eventually followed by a “reform of the reforms” in response to real or imagined problems that emerged over time. This diversity in experience is valuable since it provides an opportunity to better understand how different approaches to restructuring, competition and regulatory reform affect sector performance. However, experience drawn from different countries also needs to be utilized with some care. The objective levels of performance in different countries, in terms of costs, productivity, price levels and price structures, service quality, etc., varied widely from country to country when the reforms were implemented. The effects of the reforms on these and other performance indicia are likely to vary as well, reflecting where the sectors started as well as the nature of the reform initiatives themselves. The countries covered in the chapters in this book also vary widely in the basic legal and political infrastructure upon which the reforms had to proceed. This institutional infrastructure includes attributes, such as laws governing contracts and property rights, the powers and independence of the judicial system, experience with and effective authority of independent regulatory agencies, cost accounting, auditing and performance measurement systems, state/province versus federal authority and variations in the faith in and commitment of policymakers to competitive markets. It is particularly dangerous to draw inferences from the experience in developed countries for the design of reform policies in developing countries without carefully taking into account the differences in starting points and institutional infrastructures. The process of continuing “reform of the reforms” may provide additional insights into how different structural and regulatory mechanisms affect performance. However, continuing reform itself – an absence of a stable institutional environment for the sector – may also have adverse affects on sector performance, especially on investment incentives.

Why Restructuring, Regulatory Reform and Competition? In order to evaluate the performance of electricity sector reforms we must articulate clearly what the goals of the reforms are meant to be. I realize that some reform proponents believe that competition per se is a primary goal. Most policymakers have more precise performance goals in mind. And it is against these goals that the effects of the reforms should be measured. Electricity sectors almost everywhere on earth evolved with (primarily) vertically integrated geographic monopolies that were either state-owned or privately owned and subject to price and entry regulation as natural monopolies. The primary components of electricity 3

Although the electricity sectors in most countries started with private firms, state ownership typically came later, often after World War II. There are useful lessons that may be drawn from a careful consideration of the factors that led to a transition from private to state-owned firms in the mid-20th century, but that is not the focus of this book.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

3

supply – generation, transmission, distribution and retail supply – were integrated within individual electric utilities. These firms in turn had de facto exclusive franchises to supply electricity to residential, commercial and industrial retail consumers within a defined geographic area. The performance of these regulated monopolies varied widely across the countries. In many developing countries, the sectors were characterized by low labor productivity, poor service quality, high system losses, inadequate investment in power supply facilities, unavailability of service to large portions of the population and prices that were too low to cover costs and support new investment (Besant-Jones, 1993; World Bank, 1994; Bacon and Besant-Jones, 2001). Industrial customers sometimes had to respond to frequent system outages by building their own isolated generating facilities, increasing their costs of doing business. Sector performance in developed countries was generally much better (Joskow, 1997), but high operating costs, construction cost overruns on new facilities, costly programs driven by political pressures, wide variations in performance across firms with similar supply opportunities, and high retail prices required to cover these costs stimulated pressures for changes that would reduce costs and retail prices (Joskow, 1998, 2000a). The overriding reform goal has been to create new governance arrangements for the electricity sector that provide long-term benefits to consumers. These benefits are to be realized by relying on competitive wholesale markets for power to provide better incentives for controlling construction and operating costs of new and existing generating capacity, to encourage innovation in power supply technologies, and to shift the risks of technology choice, construction cost and operating “mistakes” to suppliers and away from consumers. Retail competition, or “customer choice,” is supposed to allow consumers to choose the retail power supplier offering the price/service quality combination that best meet their needs. Competing retail suppliers were then expected to provide an enhanced array of retail service products, risk management, demand management and new opportunities for service quality differentiation to better match individual consumer preferences. They also increase competition on the buying side in the wholesale market. It has also been widely recognized that significant portions of the total costs of electricity supply, distribution and transmission, would continue to be regulated. Accordingly, reforms to traditional regulatory arrangements governing the distribution and transmission networks have generally been viewed as an important complement to the introduction of wholesale and retail competition to supply consumer energy needs. Privatization of distribution and transmission companies combined with the application of PBR regulation imposes hard budget constraints on regulated network firms and provides better incentives for them to reduce costs and improve service quality (Beesley and Littlechild, 1989; Joskow, 2005e). In addition, the efficiency of competitive wholesale and retail markets depends on a wellfunctioning supporting transmission and distribution network infrastructure. These goals are to be achieved in a way that is consistent with environmental laws, regulations and policy goals. In many countries, the electricity sector is a major producer of conventional air pollutants (SO2, NOx, particulates) and CO2 emissions. Generating, distribution and transmission facilities can have significant impacts on land use and are not perceived as attractive neighbors by most individuals. Generating plants use large amounts of water for cooling and the release of warm water can have ecologic impacts. They also produce significant amounts of toxic waste which must be disposed of properly. Moreover, the electric power industry has been a target of environmental regulations (and related regulations governing energy efficiency and renewable energy programs) at levels that are greater than the sector’s relative impact on the environment, compared to the industrial sector, for example. I believe that this is the case because the electricity sector has been composed of regulated vertically integrated monopolies, perceived to be immune from competition from other states, provinces

4

Electricity Market Reform

and countries, and the sector provided an attractive target for “taxation by regulation,” allowing politicians to bury the costs of these programs in regulated electric power prices. These traditional approaches to use the electric power sector to pursue environmental policy agendas may not be compatible with the successful development of competitive markets. Alternative approaches that are more compatible with competitive markets may be necessary.

Textbook Architecture For Restructuring and Competition While a number of variations are potentially available (Hunt, 2002; Joskow, 2000a, 2005a), it is my view that the “textbook” architecture for restructuring, regulatory reform and the development of competitive markets for power involves several key components: 1. Privatization of state-owned utilities to create higher-power incentives for performance improvements and to make it more difficult for the state to use these enterprises to pursue costly political agendas. The components of these political agendas have included the use of state-owned monopolies for patronage employment, macroeconomic and redistributive policies, to favor domestic suppliers of fuel and equipment, and to funnel revenue to government budgets outside of the tax system. 2. Vertical separation of competitive segments (e.g. generation, marketing and retail supply) from regulated segments (distribution, transmission, system operations) either structurally (through divestiture) or functionally (with internal “Chinese” walls or “ring fencing” separating affiliates within the same corporation). These changes are thought to be necessary to guard against cross-subsidization of competitive businesses from regulated businesses and discriminatory policies affecting access to distribution and transmission networks upon which all competitive suppliers depend. 3. Horizontal restructuring of the generation segment, to create an adequate number of competing generators to mitigate market power and to ensure that wholesale markets yield reasonably competitive performance results. 4. Horizontal integration of transmission and network operations to encompass the geographic expanse of “natural” wholesale markets and the designation of a single independent system operator to manage the operation of the network, to schedule generation to meet demand and to maintain the physical parameters of the network (frequency, voltage, stability). These structural changes are necessary to provide an efficient platform for wholesale and retail competition to proceed consistent with the physical constraints and operating protocols that must govern the operation of any electric power network that meets standard reliability criteria. Horizontal integration eliminates inefficient institutional seams between physically synchronized networks, allows for more effective use of network capacity, expands the geographic expanse of competition, reduces distortions caused by inefficient transmission prices and supports the operation of wholesale markets with a minimum of intervention by system operators or regulators. An independent system operator (whether in the form a system operator with responsibility only for balancing supply and demand in real time consistent with the network’s topology and reliability criteria or a “Transco” that owns and operates the network’s transmission facilities as well) is needed to guard against discrimination that might arise if there is common ownership between markets participants who use the network and the entity that operates the network and supporting market mechanisms. 5. The creation of voluntary public wholesale spot energy and operating reserve market institutions to support requirements for real time balancing of supply and demand for

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

6.

7.

8.

9.

10.

5

electric energy, to allocate scarce network transmission capacity, to respond quickly and effectively to unplanned outages of transmission or generating facilities consistent with the need to maintain network voltage, frequency and stability parameters within narrow limits, and to facilitate economical trading opportunities among suppliers and between buyers and sellers. Physical (or near physical) spot wholesale power markets do not design themselves. They must be designed as part of the restructuring process. There are differences in opinion about what the most effective design elements are, but there should be no difference of opinion about the need for (at least) public balancing markets for energy and ancillary network support services that cover a geographic region that encompasses as large a fraction as possible of the energy trading that can take place using the network. The design of public spot energy and ancillary services markets should be compatible with the evolution of private markets for bilateral forward contracts for energy and associated derivatives, including instruments that can be relied upon to hedge basis risk associated with transmission congestion, power exchanges and other institutions to facilitate financial arrangements between buyers and sellers. The application of regulatory rules and supporting network institutions to promote efficient access to the transmission network by wholesale buyers and sellers in order to facilitate efficient competitive production and exchange, including mechanisms efficiently to allocate scarce transmission capacity among competing network users, and to provide for efficient siting and interconnection of new generating facilities. The unbundling of retail tariffs to separate prices for retail power supplies and associated customer services to be supplied competitively from the regulated “delivery” charges for using distribution and transmission networks that would continue (primarily) to be provided by regulated monopolies. This makes it possible for retail consumers eligible to choose their power suppliers competitively to purchase their power supplies from competing retail suppliers without having to overcome barriers caused by behavior of incumbents that may have the effect of increasing entry barriers for independent competitive suppliers. Where policymakers have determined that retail competition will not be available (e.g. for domestic and small commercial customers), distribution companies or alternative designated suppliers would have the responsibility to supply these customers by purchasing power in competitive wholesale markets or, if they choose, to build their own generating facilities to provide power supplies. However, in the latter case the associated charges for power would be subject to wholesale market-based regulatory benchmarks, primarily competitive procurement processes. The creation of independent regulatory agencies with good information about the costs, service quality and comparative performance of the firms supplying regulated network services, the authority to enforce regulatory requirements, and an expert staff to use this information and authority to regulate effectively the prices charged by distribution and transmission companies, and the terms and conditions of access to these networks by wholesale and retail suppliers of power, are also important but underappreciated components of successful reforms. Regulators should rely on well-designed PBR mechanisms that meet budget balance, rent extraction and efficiency criteria, given the information available to them (Joskow, 1998), and must create a stable and credible regulatory environment that will support the attraction of the capital needed to improve the performance and expand the regulated network platforms upon which competition depends. Transition mechanisms that are compatible with the development of well-functioning competitive markets need to be built into the reform program. If there is anything to be learned from the chapters in this book it is that there will be problems that need to be addressed as the reforms proceed. It makes sense to anticipate them and to build effective

6

Electricity Market Reform transition mechanisms into the reform process. These include forward contracts between generators and distributors with continuing retail customer service obligations, retail default service pricing arrangements available to retail customers before they switch to retail suppliers, and market monitoring and mitigation mechanisms. The challenge is to identify transition mechanisms that are temporary and actually support a transition rather than mechanisms that act as barriers to the development of competitive markets. It must also be recognized that it is much easier to smooth the transition when there is excess generating capacity and little network congestion than when the system has tight supplies, a lot of congestion that creates small “load pockets” and requires additional investment in generation and transmission capacity quickly to meet demand reliably.

As I read the chapters in this book it is quite clear that a few countries/states/provinces followed this textbook restructuring architecture, others embraced several of its key components, while still others took a minimalist approach to restructuring to start with and then made sequential reforms in response to (often bad) experience. As I will discuss further below, it is fairly clear that electricity sector reforms designed to create competitive wholesale and retail markets yield better performance if they follow the major elements of this textbook restructuring program than if they do not. Performance Assessments There have been few comprehensive “social cost benefit” assessments of the effects of electricity restructuring in specific countries. Newbery and Pollitt’s (1997) analysis of the welfare consequences of reforms in the UK is an exception, though it covers a period that precedes changes in the ownership concentration of generation and changes in wholesale market institutions that increased competition in the wholesale market and led to lower price–cost margins. There has been much more work on individual segments of the industry affected by the restructuring programs in particular countries (e.g. labor productivity in generation and distribution; integration of wholesale markets, investment in generation) and more of what I would call studies that examine “fragments of evidence” associated with the performance of specific segments of the sector. Most of the chapters in this volume fall in the latter two categories. Nevertheless, they convey useful information and exhibit some common themes. One of the challenges that must be confronted in doing a performance assessment is to choose a suitable counterfactual benchmark for comparison purposes. That is, we need to measure various performance metrics and compare them with what these metrics would have been if the reforms had not been made at all or if they had been made differently. The easiest approach is to examine changes in performance over time using time series data and to attribute improvements or deteriorations in performance “before and after” the reforms were implemented to the reforms themselves. This type of analysis is always based on explicit or implicit assumptions about how various exogenous variables affect performance over time and whether or not they have been controlled for adequately. So, for example, in examining the behavior of wholesale electricity prices, controlling for changes in the prices of fuels used to generate electricity over time would be important.4 Another approach is to compare the performance of electricity sectors in countries or states/provinces that have implemented reforms with those that have continued with the traditional industry structure and regulatory arrangements. This kind of approach has 4

Although the fuel price itself may be affected by the restructuring program as for coal prices in Britain before and after restructuring.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

7

been utilized extensively in the US to compare performance of various regulatory and deregulatory policy initiatives in different states that have implemented particular policies in different ways (Joskow and Rose, 1989). In a recent paper (Joskow, 2006) I have attempted to make a preliminary assessment of the effects of wholesale and retail competition programs on retail prices using cross-state data over time. This kind of cross-sectional approach has not yet been used extensively or systematically to evaluate electricity sector reforms and the chapters in this book provide a vehicle for doing this kind of cross-sectional comparison using case studies for different countries. A third approach to evaluate the performance of regulatory reform and deregulation initiatives is a “structural” simulation approach (Joskow, 2005c). Here one uses historical experience to construct a structural simulation model for the behavior and performance of the industry under different institutional arrangements. The model is then used to simulate what key performance metrics would have been if historical institutional arrangements had continued per the status quo. The simulated performance can then be compared to actual performance. With a good structural model one can also simulate how policy changes that are different from those that were implemented would have affected performance. This kind of approach has been applied to electricity restructuring policies primarily to evaluate wholesale market power issues. Newbery and Pollitt (1997) produce a counterfactual for England and Wales based on various assumptions about the incumbent state-owned generation and transmission company’s investment plans and performance trends. Green and Newbery (1992) develop a simulation model based on a particular model of imperfect competition that is parameterized to match the supply and demand attributes of the electricity sector in England and Wales in 1990. The model is used to simulate wholesale prices, generator profits, entry, consumer and producer welfare under different market structures. Wolfram (1998, 1999) effectively tests the predictions of that model using ex post data on actual prices, costs and bidding behavior. A related competitive benchmarking approach has also been applied to the evaluation of wholesale pricing behavior in California during the first few months of the so-called California electricity crises in 2000–2001 (Borenstein et al., 2002; Joskow and Kahn, 2002). All three of these approaches can provide useful insights into the effects of policy reforms on various performance indicia. However, in each case it is important to adopt what Oliver Williamson (1985) refers to as a comparative governance approach to the evaluation of the performance of alternative institutional arrangements for any industry. It has two components: (a) performance assessments must recognize that observed performance should be compared with performance under a clearly defined alternative set of institutional arrangements and (b) “ideal” textbook performance that we associate, for example, with perfectly competitive markets, is never achievable in reality. Policymakers should be looking for the best that they can do in an imperfect world.

Lessons Learned 1. Electricity sector reforms have significant potential benefits but also carry the risk of significant potential costs if the reforms are implemented incompletely or incorrectly: Based on the analyses presented in the chapters in this book, as well as other studies of electricity sector reforms and their performance, I believe that it is fair to say that when electricity restructuring and competition programs are designed and implemented well, electricity sector performance can be expected to improve significantly compared to either state-owned or private regulated vertically integrated monopolies. Note that this conclusion is not inconsistent with a finding that there are some regulated vertically integrated monopolies that perform quite

8

Electricity Market Reform

well and that, in such cases, the kinds of comprehensive reforms reflected in the textbook model might have little positive effect on performance. And if the reforms are poorly or incompletely implemented, could even lead to a deterioration in performance. Rather, it is a statement about what expectations policymakers, faced with imperfect and asymmetric information about the performance of the regulated sector, should have in the typical cases. These expectations can and should be refined by policymakers by using whatever information they have available about the pre-reform levels of performance and how far their sector is from the performance frontier. As I indicated earlier, the pre-reform levels of performance varied widely across countries and even across firms in the same country. The scope for improvements from reform initiatives necessarily vary as well. Policymakers should take the starting point into account because, as we shall see, there is always the chance that design or implementation flaws in the reforms can lead to performance problems of various kinds. Several chapters in this volume along with other research make it clear that successful implementation of liberalization reforms is not easy and that there is a risk that costly performance problems may emerge. California is the textbook case of reforms gone bad, though it is not at all clear that the right lessons have been learned from that experience. I will discuss the California case in more detail presently. As described in the chapters examining the reforms in Ontario and Brazil, they have also experienced very significant problems. Wholesale markets with good performance attributes have been slow to emerge in some countries. The promised benefits of retail competition for residential and small industrial customers have been slow to be realized in many countries. The mobilization of adequate investment to expand generation, transmission and distribution capacity has been a problem in many of the countries that have implemented reforms. Getting the reforms right at the outset is very important. 2. The textbook model of restructuring, regulatory reform and market design is a sound guide for successful reform: The use of the phrase “deregulation” to characterize the attributes of the most successful electricity sector reform programs is misleading. This is not the trucking industry and the traditional industry structure based on vertically integrated regulated monopolies is not conducive to simple “deregulation” without supporting structural, regulatory and market design reforms (Joskow and Schmalensee, 1983)! Restructuring, regulatory reform, wholesale and retail market design, and deregulation of competitive wholesale and retail segments go together. The most successful reform programs discussed in this volume and elsewhere have followed the “textbook model” outlined earlier fairly closely: privatization of state-owned enterprises, vertical and horizontal restructuring to facilitate competition and mitigate potential self-dealing and cross-subsidization problems, PBR regulation applied to the regulated transmission and distribution segments, good wholesale market designs that facilitate efficient competition among existing generators, competitive entry of new generators, and retail competition, at least for industrial customers. In my view, the gold standard for electricity sector reform is England and Wales (as discussed in David Newbery’s chapter). The reforms followed the basic architecture of the textbook model and have led to significant performance improvements in many dimensions. This is not to say that everything worked perfectly. Clearly, the decision to create only three generating companies, two of which set the clearing price in the wholesale market in almost all hours, led to significant market power problems that persisted for several years. Not only were wholesale prices too high, but there was probably an inefficiently high level of entry of new gas-fired combined-cycle gas turbine (CCGTs) technologies. Congestion on the transmission network made some generators “must run,” creating an additional “locational” market power problem. However, a combination of entry of new generators, divestitures of existing generating plants by incumbent suppliers, and transmission investments has made

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

9

the wholesale market structurally more competitive over time. Price–cost margins eventually fell dramatically and there is a lively debate about whether it was the reduction in seller concentration or the introduction of the New Electricity Trading Arrangements (NETA) to replace the pool that is the cause of the reduction in market power observed in the last few years (Evans and Green, 2005). Privatization and the application of high-powered regulatory mechanisms have led to improvements in labor productivity and service quality in electric distribution systems in England and Wales as well (Domah and Pollitt, 2001). The application of incentive regulation mechanisms to the independent transmission company also led to a dramatic reduction in the costs of managing network congestion and the costs of balancing the system and maintaining network reliability. During the 1990s there was substantial entry of new generating capacity, largely replacing existing generating capacity (that eventually retired), rather than to meet a need for new capacity to meet growing peak demand. The retail competition program in England and Wales has been reasonably successful, though there continue to be debates about whether the benefits of extending retail competition to domestic (residential) customers was worth the costs (Newbery chapter on Britain, Green and McDaniel (1998) and Salies and Waddams Price (2004)). England and Wales is not the only country that has followed the textbook model. Argentina followed most features of the basic textbook model and, prior to the country’s macroeconomic collapse, currency crisis, and rejection of contractual and regulatory commitments in 2002, experienced excellent performance. Argentina experienced significant improvements in the performance of the existing fleet of generating plants, significant investment in new generating capacity, and improvements in productivity and a reduction in losses (physical and due to thefts of service) on the distribution networks (Argentina/Columbia chapter, Estache and Rodriguez-Pardina (1998); Bacon and Besant-Jones (2001); Rudnick and Zolezzi (2001); Pollitt (2004)). Unlike the case in England and Wales, Argentina made a serious effort at the outset to create a generation sector that was structurally competitive and there is little if any evidence of market power in the wholesale market there. These improvements in performance indicia were realized despite (or perhaps partially because of) the fact that Argentina did not have a real unregulated spot market for electricity. Following the model established in Chile, Argentina’s so-called spot market was structured as a securityconstrained marginal cost-based power pool in which the clearing price is determined mechanically by the marginal cost of the generator that clears the market in an efficient costbased merit order dispatch. This mechanism effectively caps prices in the spot market at very low levels (about $150/MWh) under scarcity conditions. However, the spot market revenues are supplemented by revenues from a capacity payment mechanism. As discussed by Adib and Zarnikau in their excellent chapter, Texas also took a comprehensive approach to restructuring, regulatory reform and market design that followed many of the basic attributes of the textbook model. However, rather than adopting a poolbased wholesale market as in the UK and Argentina, Texas took an approach to wholesale market design that relied as much on bilateral contracts and as little on organized public markets operated by the ISO as possible. Texas also endeavored to implement structural remedies (i.e. generation divestiture) to respond to concerns about market power. However, congestion management and associated market power issues have been an area of concern in Texas and it appear that Texas will be moving to a nodal pricing model more like those operating in the Northeast in the future. Texas also adopted an approach to retail competition that is similar to that adopted in the UK, except retail competition was opened to all classes of customers from the beginning. At least in terms of switching behavior, Texas has the most successful retail competition program

10

Electricity Market Reform

in the US, especially for smaller customers. However, I question the estimates of customer savings achieved by retail competition based on comparisons between the prices available from competitive retailers and the “price to beat.” I do not think that the “price to beat” is the appropriate benchmark against which to measure actual or potential customer saving relative to what prices would have been under regulation, especially with extremely high natural gas prices. Indeed, there is evidence that retail prices in Texas, especially for customers who have not switched, are higher than they would have been during the short period of time retail competition has been in effect in Texas (Joskow, 2006). If regulation had continued, customers would have been partially hedged against changes in natural gas prices as the costs of nuclear and coal-fired capacity would have been averaged in with the cost of gas-fired capacity in determining regulated retail prices. The price to beat does not reflect this hedge. Nor do competitive wholesale market prices in markets that clear with gas-fired capacity during a large fraction of the hours in the year reflect this hedge (as in New England). New Zealand, portions of Australia and the Nordic countries adopted many of the key components of the textbook model and have had reasonably successful reform programs, though retail competition opportunities vary between these countries. Australia, the Nordic countries, Ontario, Australia and Brazil have proceeded with their reforms without fully privatizing the generation segment of the sector. The continued mix of public and private generating companies raises some interesting issues for short-run market performance and longer-run investment incentives that I will return to presently. Chile is often identified as the first country to adopt the textbook electricity sector reform model. While I believe that the Chilean reforms have led to large efficiency improvements compared to what proceeded them, and that there is much to be proud of in the reforms that were made there in the beginning of 1980s, the Chilean system has involved less restructuring, less competition and more regulation than first meets the eye (Joskow, 2000b). While Chile theoretically separated generation, transmission and distribution, for many years a single holding company owned the largest generating company, the primary distribution company in the Santiago region and the primary transmission company serving the largest region of the country.5 The generation segment is not structurally competitive and there is significant potential market power that would be exercised if Chile really had a competitive wholesale market, but it does not. What is generally referred to as a spot market in Chile is not really a market in the same sense as are the spot markets for energy in PJM, Alberta, England and Wales, or Norway. Indeed, it is little different from the pre-restructuring centrally dispatched power pools that existed in the US for the last few decades. Generators are dispatched based on estimates of their marginal production costs and the marginal cost of the last supply unit called to meet demand determines the market clearing price. Network congestion and constraints are centrally managed by the system operator (the CEDEC in Chile) in conjunction with the “least-cost” dispatch of generators. Generators have incentives to keep their costs low and their availability high under this system. However, the “spot market” does not yield prices that are high enough when generating capacity is fully utilized to balance supply and demand (the effective price cap is on the order of $150/MWh) and, as a result, scarcity rents are too low to attract adequate investment in generation to the market. Nor has there been a market mechanism to pay for operating reserves and other ancillary services. Accordingly, it should not be surprising to find that the incentives to invest

5

The transmission company Transener was sold to an affiliate of Hydro Quebec by Endesa after its acquisition of the Enersis holding company which had owned the largest distribution, transmission and generating companies in the SIC region of Chile.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

11

in new generating capacity have been inadequate. Moreover, the government’s decision to go out to tender for additional generating capacity is not an attractive solution to the investment incentive problems caused by wholesale and retail market imperfections. On the retail competition front, large industrial customers in Chile were theoretically free to contract directly with generators for their supplies (though they are not permitted to buy directly from the “spot market”), but as a practical matter only the very largest customers which could connect directly to the high-voltage transmission system have this opportunity. I understand that the distribution company serving Santiago had many customers that theoretically could contract directly with generators, but very few actually took advantage of this opportunity. The reasons are (a) there was no unbundled delivery tariff available to customers that separated delivery charges from generation charges and (b) generators were reluctant to steal the distribution company’s retail customers since the distribution company itself was a major contract purchaser of the generators’ wholesale power supplies. Under the Chilean model, distributors were supposed to enter into contracts with generators to meet their load obligations, but the prices in these contracts are regulated based on forecasts of nodal prices in the “wholesale market” and the associated costs are, in turn, passed through to retail consumers. The nodal prices are theoretically collared by the “free market” prices paid by large industrial customers – they must be in a ⫾10% band of negotiated contracts between generators and large customers, but, as I have just noted, the competition in the free market is more limited than first meets the eye.6 Accordingly, there is not really a competitive wholesale contract market either. Theoretically there has been free entry of new generators in Chile for many years, but until a new law was passed in 2004 there was no open access transmission tariff,7 no obligation placed on transmission companies to plan for or build transmission capacity in advance, and the major transmission company was owned by the major generator and they must, by necessity, interact closely with one another. There has been little entry of new generating companies in Chile, though existing generating companies have expanded generating capacity significantly over time.8 Nevertheless, creating adequate investment incentives for new generating capacity is a continuing issue in Chile. So, whatever the success that the Chilean reforms achieved, they are not primarily the result of vibrant unregulated competitive wholesale or retail markets for electricity or real vertical and horizontal restructuring. Privatization, incentive regulation, a simulated competitive spot market, contractual obligations placed on distribution companies and free entry by incumbent suppliers in response largely to administratively determined generation prices have all contributed to the performance improvements. California and many of the Northeastern US states appear to have adopted many of the components of the textbook model as well. Yet California is often put forward as the textbook case of “deregulation” gone bad. The California restructuring and competition program (but not the T&D regulatory framework) were heavily influenced by the earlier reforms in England and Wales. Several of the commissioners of the California Public Utilities Commission (CPUC) visited England in early 2004 and were very impressed with what they saw there. The initial reform proposals contained in the so-called “blue book” included many of the features of the reform program in England and Wales. And, although disputes about wholesale

6

Prices negotiated in the free market are confidential and I have seen no analysis that indicates whether the negotiated contract prices are a binding constraint and, if they are, how large is their effect on nodal prices. 7 Such as provided for by Order 888 in the US or the Grid Code in England and Wales. 8 The relatively recent availability of natural gas in Chile may make entry of new suppliers and competition from cogeneration easier in the future.

12

Electricity Market Reform

and retail market design led eventually to a reform program that departed from several aspects of the textbook model, it still retained many of its basic features. California’s utilities were effectively required to divest their fossil-fueled generating plants in a way that led to a significant reduction in horizontal concentration and the instant creation of a large independent power sector. California’s generating sector had relatively low levels of concentration after the incumbent utilities divested their gas/oil-fired generating capacity, especially when California’s substantial interconnector capacity is taken into account. Moreover, the financial and regulatory commitments placed on the incumbent utilities during a transition period destroyed any incentive they might have had to exercise market power using their remaining generating plants. An independent system operator was created to consolidate the control areas of California’s three investor-owned utilities and to operate a single control area transmission network and manage congestion. Public day-ahead spot markets for power (the PX), a separate real time balancing market for energy (operated by the ISO) and ancillary service markets were created (operated by the ISO), and there was some integration between these markets and the allocation of scarce transmission capacity yielding a set of zonal prices (though intra-zonal congestion was not priced transparently and the associated costs were included in an uplift charge). All customers were made immediately eligible for retail competition or so-called “customer choice.” While there were a number of problems that emerged in the wholesale markets in California after they began operating in 1998, wholesale market prices between April 1998 and April 2000 were roughly equal to what had been predicted (3 cents/kWh) prior to the reforms and well below the embedded cost of regulated generation service (about 6.5 cents/kWh). The meltdown began in June 2000 and continued for about a year. What was the source of the problems? Interestingly, the New England states, New York, New Jersey and Pennsylvania had implemented very similar reforms at about the same time and experienced some of the same exogenous shocks to demand and fuel prices in 2000 and 2001. Yet they did not experience the same system meltdown as did California. So, there is something to learn as well from comparing some of the more detailed aspects of the reforms in California with those in these other US states. Many explanations have been advanced to explain what happened in California. One interpretation of what transpired and why can be found in James Sweeney’s chapter in this volume. My views, written at about the time the crisis was winding down and before the Enron and other marketers’ tapes were released, can be found in Joskow (2001). The most frequent explanation that I hear is that there was a shortage of generating capacity in California and that this shortage was a result of poor investment incentives inherent in California’s wholesale market design. This is nonsense. There was little investment in generating capacity anywhere in the US during the time period when the California reforms were being designed and implemented (1994–1998). This is because there was excess capacity in most regions of the US during the early 1990s. Uncertainties about the future path of restructuring, regulatory and competitive reforms that began to be discussed seriously at this time was also a deterrent to potential investors waiting until the rules of the game were specified more clearly. Indeed at the time of the crisis there was a long queue of developers that had applied for permits to build new generating plants in California after the market opened in April 1998. It is unrealistic to expect that even under the best of circumstances any significant amount of new generating capacity could have come out of the construction pipeline in 2 years. Moreover, California is a summer peaking system. The biggest problems, in terms of high prices, operating reserve emergencies and rolling blackouts did not occur until the winter of 2000–2001. The problem was not that there was inadequate physical generating capacity in place, but rather that a large fraction of the existing generating capacity was not available

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

13

to generate electricity. The best that can be said from a resource adequacy perspective is that a lot of generating capacity that should have been available to meet demand had broken down and was not capable of producing electricity. The “shortage” of generating capacity may perhaps be explained by older plants breaking down and by their owners’ reluctance to supply when it became unclear about January 2001 whether or not they would be paid. However, there is also abundant evidence that some suppliers exploited opportunities to engage in strategic behavior to jack up market prices. At least in the summer of 2000, some generators were taking advantage of a tight supply situation to exercise market power (Borenstein et al., 2002; Joskow and Kahn, 2002). The tapes of the conversations of traders for Enron and other companies that subsequently were released make it clear that they saw and took advantage of opportunities to withhold supplies and increase market prices during the crisis. It is true that California’s wholesale market would have been stressed during the second half of 2000 even if there had been no market power problems. Demand was unusually high throughout the Western Interconnection, natural gas prices and NOx permit prices rose significantly. However, even after taking account of these factors it is hard to explain what happened during the second half of 2000 only as the result of the interplay of supply and demand in a competitive market. It is interesting to note that the Northeast experienced similar exogenous shocks to those experienced in California at this time. Little new generating capacity had come out of the construction pipeline yet and supplies were tight during the summers of 2000 and 2001. While wholesale prices rose, there was no meltdown similar to what took place in California. What made the situation in the Northeast different from California? It wasn’t a more competitive generation market structure since the Northeastern markets had more concentrated generating sectors than California. However, unlike the case in California, in the Northeast when utilities divested their generating plants and took on regulated retail default service obligations during a transition period they also entered into forward contract obligations to match their retail supply obligations and were largely (though not completely) hedged against large movements in wholesale prices. The utilities in California were not permitted to hedge fully their retail supply obligations, though contrary to the conventional wisdom, the two largest California utilities were partially hedged as a result of the nuclear, hydro, coal and existing power supply contracts they retained. In addition to providing a hedge, the “vesting contracts” in the Northeast also created incentives for the generator counterparties to these contracts to supply at least enough to meet their supply commitments and significantly mitigated any potential market power that they might have had.9 Moreover, most of the utilities in the Northeast were permitted to recover (with a lag) wholesale power costs that exceed the regulated default service prices that had been negotiated with their regulators. As wholesale market prices rose along with natural gas prices, distribution companies could book reasonable wholesale market purchase costs and recover them later through their delivery or retail default service prices. In California, the CPUC initially took the 9

These transition hedging arrangements have or soon will expire and the expected transition to competitive supply of retail customers has been quite slow. This means that distribution companies are increasingly going to the short-term wholesale market to arrange for wholesale power supplies to meet the loads of their customers who have not switched to competitive supplies. As this is written, forward wholesale market prices are significantly greater than the prices reflected in the expired(ing) transition contracts, largely as a result of large increases in natural gas prices, and retail customers relying on default service will soon see a large increase in their electricity rates. This is likely to increase the influence of those who argue that the pro-competition reforms were a mistake.

14

Electricity Market Reform

position that retail prices were fixed during the transition period and that no additional cost recovery was allowed. This decision first created a payment crisis by January 2001 and then led to the actual bankruptcy of PG&E and the de facto bankruptcy of SCE. In my view, if California had implemented similar transition arrangements to those implemented in the Northeast, in particular if the California utilities had more completely hedged their retail supply obligations with forward contracts and had the opportunity to recover from retail customers reasonable costs of the power they purchased in wholesale markets, there would have been no California electricity crisis. This is not to say that deficiencies in the design of California’s wholesale markets would not have led to inefficiencies that would have driven up wholesale power costs to some degree. Rather, there would not have been a sudden financial collapse and California would have had time to improve its wholesale market and transmission institutions as in the Northeast. Instead, California responded to the crisis with costly long-term contracts negotiated by the state, long-term procurement obligations, a freeze on retail competition and a strange mix of regulatory obligations and competitive markets that does not bode well for the future. 3. Departing significantly from the textbook model of restructuring, competitive market institutions and regulatory reform is likely to lead to performance problems: In the reform cases discussed above and others covered in the book, restructuring, regulatory reform and the creation of wholesale and retail market institutions proceeded more or less in parallel. There was a clear recognition that a viable retail market depended on the existence of a robust wholesale market and that a robust wholesale market depended on horizontal and vertical restructuring and regulatory reform at the T&D levels. This is not how the reforms proceeded in much of continental Europe (Spain and the Netherlands being the primary exceptions), in Japan, and in large portions of the US. The reform process stimulated by European Union (EU) directives was in my view especially strange, but perhaps politically quite astute. The initial focus of the EU reforms was on “market opening” for retail customers. That is, the focus was on retail competition. This focus ignores the fact that “market opening” alone will not lead to meaningful retail competition in the absence of appropriate wholesale market and network access and pricing institutions. Retail customers may be given the freedom to shop around for their power needs, but unless they can obtain delivery services on reasonable terms and conditions, and there is a well-functioning competitive wholesale market where they or their agents can shop, there will be no meaningful opportunity to take advantage of this freedom. Accordingly, it should not be surprising that “market opening” for retail competition alone doesn’t lead to much in the way of meaningful competition. I view the slow pace of development of competition in many of the countries in continental Europe as being largely attributable to their failure to restructure vertically and horizontally and to create the necessary network access, pricing and wholesale market institutions to create a robust wholesale market. The situation described in the chapter on Germany provides the starkest example of how retail competition without restructuring and the creation of competitive market and supporting regulatory institutions leads to performance problems. The German electric power system continues to be dominated by vertically integrated utilities with interests in generation, transmission and distribution. They control the operation of the transmission networks, which are operated as separate control and balancing areas rather than as a single balancing area as in other European countries. There is no independent system operator. Generation ownership is fairly concentrated. Until recently, there was no regulator to determine network costs and prices or to enforce unbundling rules necessary to support retail and wholesale competition, however, the anti cartel office followed the transformation of the industry and made a number of important decisions. Competition developed quite dynamically in the

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

15

wholesale market based on an agreement between large consumers and suppliers and supported by a power exchange. Without a clear regulatory framework it should not be surprising that meaningful competitive retail markets have been slow to emerge in Germany. The modest reform program laid out in the chapter on Japan suggests that Japan is on a similar path. Whether it is by design or accident, however, the EU’s focus on market opening for retail consumers has led it to look more closely at supporting reforms upstream at the wholesale and transmission levels as time has passed. EU countries are now creating independent system operators and transmission entities, sometimes relying on ownership separation and sometimes on functional separation or ring fencing. Germany has been forced to create a regulator to regulate (at least) network charges and unbundling protocols. While the EU and other panEuropean institutions have focused on transmission facilities that connect individual member countries, rather than getting involved in intra-country market design or competition issues, the member countries are increasingly realizing that efficient use of interconnector capacity requires some compatibility between intra-country wholesale market designs and coordination between them (ETSO and EUREPX documents). In my view, the EU started at the wrong end of the system, but given the EU’s limited legal authority perhaps this was the only way they could get a credible reform process going in light of the opposition of many incumbents to reform. The EU and members countries are now moving back upstream to implement a variety of structural and institutional reforms that would have, ideally, been done first rather than last. The chapter on Brazil provides another perspective of how reforms can go bad. The reforms in Brazil were accompanied by a reasonably complete reform blueprint. However, the blueprint was only partially implemented and the program was overwhelmed by a water shortage that would have led to problems under any circumstances. The problems were probably worse because of the incomplete implementation of the reforms and were blamed unfairly on the reforms themselves. 4. Public spot energy and ancillary services markets should be integrated with the allocation of scarce transmission capacity. The most efficient design of spot wholesale energy markets continues to be a subject of dispute among interest groups and independent experts (Hunt, 2002; Stoft, 2002; Joskow, 2005a). Should the market be built around a pool or rely on bilateral contracts? Should there be locational pricing of energy and operating reserves? How should scarce transmission capacity be allocated? Should transmission rights be physical or financial (Hogan, 1992; Joskow and Tirole, 2000)? While there is some room for flexibility, and some of the disputes reflect the self-serving arguments of interest groups that expect to benefit from inefficient markets, I believe that the experience to date supports the desirability of several basic wholesale market design features. These basic design features include the creation of voluntary public spot markets for energy and ancillary services (day-ahead and real time balancing) that accommodate bilateral contracts and self-scheduling of generation; locational pricing reflecting the marginal cost of congestion and losses at each location; the integration of spot wholesale markets for energy with the efficient allocation of scarce transmission capacity; auctioning of (physical or financial) financial transmission rights that are simultaneously feasible under alternative system conditions to hedge congestion, serve as a basis for incentives for good performance by system operators and transmission owners, and partially to support new transmission investment;10 an active demand side that can respond to spot market price 10

The allocation of transmission rights can, however, affect the incentives of firms to exercise market power and this should be taken into account in the design of rights allocation mechanisms and restrictions on the entities that can purchase these rights (Joskow and Tirole, 2000; Gilbert et al., 2002).

16

Electricity Market Reform

signals (Borenstein et al., 2002). These are the attributes of the PJM markets, as well as those in New England, New York and the Midwest ISO in the US (Joskow, 2006). California is proposing to implement a similar “nodal price” market design and Texas is considering it. 5. Market power is a significant potential problem in electricity markets, but the cure can be worse than the disease. Try to deal with potential market power structurally ex ante rather than ex post. The potential for market power to be a particularly severe potential problem in electricity markets was recognized many years ago (Joskow and Schmalensee, 1983, Chapter 12). It arises as a consequence of transmission constraints that limit the geographic expanse of competition, generation ownership concentration within constrained import areas, the non-storability of electricity and the very low elasticity of demand for electricity (Joskow, 1997; Borenstein, 2002). Generator market power was a serious problem for several years following the launch of the privatization, restructuring and competition program in the UK (Wolfram, 1999). Concerns about market power in the US were reinforced by the events in California in 2000–2001 (Borenstein et al., 2002; Joskow and Kahn, 2002) where market power and the exploitation of market design imperfections contributed to the explosion in wholesale prices beginning in June 2000. Market power issues of various kinds have been identified by chapters in this volume and elsewhere (New Zealand, Chile, Columbia, PJM, Texas, Alberta, Brazil and some areas of Central Europe). The problems can be attributed to the interactions between the attributes of electricity networks noted above, too few competing generating companies, wholesale market design flaws, vertical integration between transmission and generation that creates the incentive and opportunity for exclusionary behavior, excessive reliance on spot markets rather than forward contracts, and limited diffusion of real time prices and associated communications and control technology that facilitates the participant of demand in wholesale spot markets. Clearly, market power is an issue that must be taken seriously. No market design will work well if there are not an adequate number of competitive suppliers of generation service or the market power of dominant firms has not been mitigated in some way (i.e. with regulated forward contracts). Ex ante structural remedies are likely to be superior to ex post behavioral remedies, but the former are difficult to implement once generators have been privatized and deregulated. As a result, market power mitigation strategies have become an important component of wholesale market reforms in many countries. In the US, FERC market monitoring and market power mitigation protocols have been a central component of all of its reform initiatives. All of the ISOs in the US have market monitoring units, wholesale price caps have been implemented and special bidding and mitigation restrictions have been placed on generators located in small geographic load pockets. These market monitoring and mitigation protocols appear to have been reasonably successful in mitigating the ability of suppliers to exercise significant market power in these situations as well. Indeed, these measures may have been too successful, constraining prices from rising to competitive levels when demand is high, capacity is fully utilized and competitive market prices should reflect scarcity values that exceed the price caps in place. Thus, these efforts to mitigate market power in the short run may create adverse generation investment incentives in the long run (Joskow and Tirole, 2005b), a subject to which I shall return presently. 6. Good transmission and distribution network regulatory institutions are important but sometimes neglected components of the reform process. Most of the chapters in this book focus on the development of wholesale and retail markets. But it is important to remember that the textbook model includes the development and application of a well-designed regulatory framework to govern the distribution and transmission networks that will continue to be subject to government regulation of prices, costs, service quality, access rules, and investment programs.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

17

These “residual” regulated segments of the electricity sector often represent a significant fraction of the total retail price for services paid for by consumers (prices for competitive plus regulated services). Moreover, the performance of the regulated segments can have important effects on the performance of the competitive segments since the regulated segments provide the infrastructure platform upon which the competitive segments rely (e.g. the electric transmission and distribution networks). Accordingly, the welfare consequences of electricity sector restructuring and competition reforms depend on the performance of both the competitive and the regulated segments of these industries. Regulatory reform focused on applying incentive or PBR regulatory mechanisms was a central feature of the liberalization program in the UK and the regulatory institutions and mechanisms that have evolved there also represent the gold standard of effective incentive or performance-based network regulation (Beesley and Littlechild, 1989; Joskow, 2005d). Privatization and the application of high-powered regulatory mechanisms has led to improvements in labor productivity and service quality in electric distribution systems in England and Wales, Argentina, Chile, Brazil, Peru, New Zealand and other countries (Newbery and Pollitt, 1997; Estache and Rodriguez-Pardina, 1998; Bacon and Besant-Jones, 2001; Domah and Pollitt, 2001; Rudnick and Zolezzi, 2001; Pollitt, 2004). Sectors experiencing physical distribution losses due to poor maintenance and antiquated equipment, as well as resulting from thefts of electric service, have generally experienced significant reductions in both types of losses. Penetration rates for the availability of electricity to the population have increased in those countries where service was not already universally available and queues for connections have been shortened. Distribution and transmission network outages have declined. Improved performance of regulated distribution (and sometimes transmission) systems has accompanied privatization and the application of high-powered PBR mechanisms almost everywhere it has been tried. The debates about alternative regulatory frameworks in Germany (as reviewed in Brauknecht and Brunekreeft’s chapter) assumed that regulators have a stark choice between applying cost-of-service regulation or incentive regulation, implying that these regulatory frameworks are substitutes. I do not think that this is a correct perspective regarding the choice of regulatory framework and associated mechanism design choices in practice. In fact, costof-service regulation and incentive regulation are complements rather than substitutes. Much of the cost information and financial analysis required to implement a cost-of-service regulatory regime is also required to implement an incentive regulation regulatory regime. Incentive regulation is an enhancement to traditional cost-of-service regulation not a replacement for it (Joskow, 2005d). Nor is pure price cap regulation optimal in theory (Schmalensee, 1989) or used in practice. Price caps with “ratchets” that reset prices every few years to reflect prevailing costs are frequently utilized. However, the reset process makes prices partially contingent on actual cost realizations over time. Price caps with ratchets are more like a sliding scale or profit sharing arrangement than the pure price caps that appear to theoretical models. Moreover, in order to reset prices to reflect costs from time to time it is necessary to have a good capital and operating cost accounting system, to measure key financial variables like the cost of equity and debt, to agree on asset valuation (rate base or regulatory asset base), and associated depreciation policies (Joskow, 2005d). This is the same kind of information that is required to implement a cost-of-service regulatory regime. That’s why cost-of-service regulation and incentive regulation are complements rather than substitutes. It is also now widely recognized that cost reduction efforts by network owners can potentially lead to a deterioration of service quality – increases in network outages, delays in service restoration, delays answering telephone inquiries. Accordingly, well-designed regulatory programs include PBR mechanisms that apply to various dimensions of service quality

18

Electricity Market Reform

(Joskow, 2005d). These mechanisms reward or penalize network companies based on their performance against pre-specified service quality benchmarks. Bertram’s chapter on New Zealand contains a comprehensive discussion of restructuring of the distribution sector and its regulation. It is one of the few papers that I have seen that examines issues associated with network asset-valuation decisions and how these decisions can affect the prices paid by consumers for distribution service (see also Bertram and Twaddle (2005) upon which this section of the chapter on New Zealand is based). When electricity sector restructuring takes place, one decision that must be made is how to value the assets of the distribution and transmission companies that will be used for regulatory purposes going forward. The typical approach has been to carry forward the existing depreciated book value of historical investments in transmission and distribution into the new regime so that the base level of distribution and transmission charges associated with the recovery of capital-related charges does not change as a consequence of the transition. Incremental investments are then accounted for more or less as they were under the old regime (as in the US and Canada) or economic/inflation accounting methods and approximations to economic depreciation are applied (as in the UK). In New Zealand, however, a decision was made to “write up” the value of distribution assets to reflect a specific measure of their (higher) replacement cost (called “optimized deprival value” (ODV)) and to use these higher valuations to set the base level of network prices. This valuation method led to higher prices and higher price–cost margins for distribution network owners. The argument for adopting this valuation approach was that this would allow prices to rise to their efficient level and provide consumers with appropriate price signals. The arguments against this revaluation were that (a) it would lead to significant price increases and unfairly burden consumers, (b) non-linear pricing could be used to restore the correct price incentives on the margin and (c) it created windfall profits for distribution network owners and undermines support for restructuring and competition. While Bertram’s discussion focuses on the effects of this asset-revaluation program on distribution service price levels in New Zealand, the research results reported in Bertram and Twaddle (2005) also demonstrate that operating costs incurred by distribution companies in New Zealand fell very significantly during the same period of time. These cost reductions appear to reflect both the consolidation of small distribution companies through mergers and the incentives for cost reduction provided by a high-power incentive scheme. Moran’s chapter on Australia identifies significant productivity improvements in distribution as well, without any apparent deterioration in network reliability. After the first few years following restructuring significant productivity improvements in both distribution and transmission were realized in the UK as well. Effective regulation of networks does not occur by accident. It requires good regulatory institutions. Regulatory institutions that are independent, are well staffed and have access to necessary information about costs, prices and service quality continue to be an important linchpin of successful electricity reform programs. Inadequate attention has been paid to creating good regulatory institutions in many countries. Germany and New Zealand’s initial decisions to proceed with a liberalization initiative without any sector regulator at all, relying instead on negotiated prices and the constraints of competition law, were clearly a mistake. I do not think that the issues are as complicated as may be suggested by Knieps’ chapter. The presence of an electricity network regulator with the proper goals and tools is a necessary component of a successful electricity sector reform program. 7. Creating a well-functioning transmission investment framework is important but continues to be a significant challenge in many countries. As wholesale markets have developed, congestion on the transmission network has not only increased but is increasingly recognized as a

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

19

significant constraint on the development of efficient competitive wholesale markets for power. In several of the countries studied in this volume, investment in transmission capacity, especially interregional transmission capacity, has not kept pace with the expansion in demand, generating capacity, or the volume of wholesale trade. In Europe and the US there has been almost no investment in interregional transmission capacity since the early 1990s. Inadequate transmission investment is identified as a problem in Brazil and in Chile as well. Texas (ERCOT) appears to have responded to intra-regional transmission congestion with new investment, but ERCOT is still effectively disconnected from the rest of North America. In addition to the effects of transmission congestion on wholesale power prices and the associated social costs of congestion, a congested transmission network makes it more challenging to achieve efficient wholesale market performance. Transmission congestion and related reliability constraints create load pockets, reducing effective competition among generators and leading policymakers to impose imperfect market power mitigation rules that create other distortions. Congestion makes it more challenging for system operators to maintain reliability using standard market mechanisms, leading them to pay specific generators significant sums to stay in the market rather than retire and to rely more on out of market (OOM) calls that depress market prices received by other suppliers. In New England, the amount of generating capacity operating subject to special reliability contracts with the ISO has increased from about 500 MW in 2002 to over 7000 MW projected for 2005 (ISO New England, 2005, p. 80), amounting to over 20% of peak demand.11 These responses to transmission congestion undermine the performance of competitive markets for energy, exacerbate the net revenue problem discussed above and lead to additional costly administrative actions to respond to market imperfections resulting from transmission congestion. In the UK and Argentina, the restructuring process included a comprehensive set of institutions and regulatory mechanisms to govern transmission operating cost and reliability, the allocation of scarce transmission capacity and approvals of transmission investment programs, as an integral aspect of the reform process. In many other countries, the regulatory framework governing transmission operation and investment was not given too much attention and allowed to evolve along with the markets. Stimulating performance improvements in the operation of transmission networks and, especially, attracting adequate investment to reduce congestion and to increase the geographic expanse of competition to reduce market power and the associated need to regulate wholesale markets to mitigate it, has been a challenge. The transmission systems that have exhibited the best performance are organized with a single independent transmission company that spans a large geographic area, integrates system dispatch, congestion management, network maintenance and investment under PBR regulation (e.g. NGC in England and Wales). Fragmented transmission ownership, separation of system operations from transmission maintenance and investment and poorly designed incentive regulation mechanisms reduce performance. Relying primarily on market-based “merchant transmission” investment, that is where new transmission investments must be fully supported by congestion rents (the difference in locational prices times the capacity of a new link) is likely to lead to inefficient investment in transmission capacity (Joskow and Tirole, 2005a). The frameworks for supporting transmission investment in many countries continue to have deficiencies. Progress is being made, however. PJM and other ISOs in the Northeastern US have adopted spot market mechanisms that integrate energy and ancillary services markets

11

FERC has ordered the ISO to replace these agreements with a locational capacity market mechanism built around an administratively determined “demand curve” for generating capacity. However, implementation has now been delayed until at least October 2006.

20

Electricity Market Reform

with the allocation of scarce transmission capacity. They have also refined their transmission planning and investment programs significantly to capture investment needs driven by both reliability and economic considerations, though their interdependence has been slow to be recognized. These processes accommodate merchant transmission investments but do not rely on them. Chile has also introduced transmission pricing reforms recently. The Nordic markets take a different approach to integrating day-ahead energy markets with the allocation of scarce transmission capacity, but a transmission investment framework appears to be much less well developed there. Mechanisms being developed through cooperative activities of European transmission and power exchange operators and regulators for integrating energy markets with the allocation of scarce inter-country transmission capacity are moving forward in Central Europe. And recent EU rules governing investment in interconnector capacity that expands transmission capacity between countries are very constructive. There is still a great deal of work to be done on the creation of effective institutional arrangements governing the organization of transmission operations, operating costs, congestion management, reliability and investment to expand capacity. 8. “Resource adequacy” is a concern of policymakers in almost every country. The jury is still out on whether and how competitive power markets can stimulate appropriate levels of investment in new generating capacity in the right places at the right times. Many policymakers are increasingly expressing concerns about “supply security” and “resource adequacy.” It is not always very clear precisely what these phrases refer to (see the excellent conceptual discussion of different dimensions of supply security in the chapter on the Nordic market by Von der Fehr, Amundsen and Bergman). One dimension of supply security relates to the operating reliability of the network as measured by involuntary losses of power – non-price rationing or controlled rolling blackouts – given the existing stock of capital on the network. Customers may experience blackouts due to failures on the distribution system, the transmission system, or due in inadequate generating capacity and price sensitive/interruptible demand to balance supply and demand in real time consistent with maintaining the physical integrity of the network. Failure to keep the system in balance can lead to cascading uncontrolled blackouts and network collapses affecting large regions (as occurred in the US and Italy in 2003). There is also a longer-run concept of “supply security” that reflects the adequacy of investments in distribution, transmission and generating capacity. Over time, investment in additional capacity should be made as long as the incremental value of the investments exceeds the incremental cost of the investment. If too little investment is made, costs and prices, including the costs associated with non-price rationing of demand and network collapses as discussed above, will be too high. Thus, long-run concepts of supply security or resource adequacy are related to short-run concepts of supply security or network reliability. I have already discussed network investment issues and now will turn to issues associated with investment in new generating capacity. Creating appropriate investment incentives for new generating capacity is perceived to be a growing problem in many countries. At first blush, this concern may be surprising since the early experience with reforms during the 1990s suggested that competitive wholesale markets could and would mobilize adequate (or more than adequate) investment in new generating capacity. Substantial amounts of capital were mobilized during the late 1990s to support construction of new efficient generating capacity in many countries that have implemented reforms. In the US, over 200,000 MW of new generating capacity went into service between 1999 and 2004, most of it merchant capacity, an increase of nearly 30% in total US generating capacity (Joskow, 2006). About 40% of the stock of generating plants in service in England and Wales was replaced with modern efficient CCGT technology between 1990

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

21

and 2002 as old coal-burning generators have been closed and expensive dirty coal plants have been displaced by cheaper and cleaner CCGT capacity. Many other countries implementing reforms during the 1990s, including Argentina, Chile and Australia, also attracted significant investment in new generating capacity (Jamasb, 2002) after the reforms were initiated. So, why are policymakers so concerned now? First, we should recognize that liberalization has evolved in much of Europe during a period when there was significant excess generating capacity, Spain and Italy being the major exceptions. Even in the UK, the quantity of generating capacity in service today is not much greater than it was in 1990, with most of the investment in generating capacity during the 1990s being stimulated by opportunities to replace the inefficient stock of old generators that the state-owned Central Electricity Generating Board (CEGB) kept in service to maximize consumption of expensive British coal, long-term contracts entered into by retail suppliers early in the UK’s liberalization program, and the high prices available in the wholesale market, influenced by the exercise of market power as already discussed. These investments were not the result of a significant need for new generating capacity to meet rapidly growing peak demand, as is the case in several of the countries discussed in this volume. Second, the environment for financing new generating investments has changed dramatically in the last few years as a result of financial problems faced by merchant trading and generating companies in Europe, the US and Latin America, as well as macroeconomic and political instability in Latin America and Asia (De Araujo, 2001; Jamasb, 2002; Joskow, 2005a). After peaking at 55,000 MW of new capacity entering service in the US 2002, only about 15 000 MW of new generating capacity entered service in 2005, most of which was built either for municipal utilities that have not been subject to restructuring and competition reforms or wind projects that benefit from special subsidies and contractual arrangements. Concerns about future incentives for investment in additional generating capacity are noted in several of the studies in this book (Chile, Brazil, Columbia, New Zealand, California, Germany, PJM). In some cases, state-owned entities have stepped in to contract for additional generating capacity (e.g. Chile, Brazil, New Zealand, Ontario, California) to mitigate resource adequacy concerns. The actions by state-owned entities to support investment in new generating capacity may have salutary short-run effects, but these actions are likely to discourage private investment in the longer run. Programs designed to stimulate investments in renewable generation (mostly wind) with special tax subsidies, contractual benefits, or mandatory purchase obligations, further complicate the investment picture for “ordinary” generating plants. Potential private investors in new generating capacity are looking for stable market rules and longer-term contractual commitments before they will commit capital for new generating facilities. Continuous market redesign, regulatory actions that limit prices, system operators’ “reliability” actions that depress market prices, and other market and regulatory imperfections are being pointed to as deterrents to private investors in unregulated generating plants. Financing investments in peaking capacity, which rely heavily on wholesale market prices creating “rents” to support fixed investment costs in a relatively small number of hours, is especially problematic. Analyses done of regional markets in the US make it fairly clear that “energy-only” markets do not produce adequate revenues to attract investment in generating capacity consistent with the reliability standards that are still applicable to them and have now become mandatory (Joskow, 2005a, 2006). The chapters on the UK, Norway and Australia are more optimistic about generation investment incentives. I am unconvinced, though I hope that they are right. The UK is in a rather unique position since the investment wave of the 1990s has left it with a relatively large stock of older mothballed-generating units that can enter and leave the market on relatively

22

Electricity Market Reform

short notice as forward (e.g. 1 year) contract prices change, providing a fairly elastic shortrun supply of “reserve” generating capacity. A stock of mothballed reserve capacity of this magnitude does not exist in most countries and investment in new capacity is forecast to be needed to meet growing demand with conventional levels of reliability. Von der Fehr, Amundsen and Bergman’s paper on Nordic markets is also optimistic, but I am not convinced by the evidence presented to support this optimism. The fact that spot market and forward prices have not justified new investment in the last several years begs the question of whether or not these prices are providing the right price signals. This is an issue that has been analyzed for all of the organized US markets and the answer is the same – wholesale energy market prices do not provide adequate price signals (Joskow, 2006). There has been little if any net increase in generating capacity in the Nordic market area for many years and reserve margins are declining. There are a few generating projects in process, including a nuclear plant in Finland developed by a public/private consortium. Maybe the stateowned generating companies will step in and fill any gaps, but whether this will be a market response or a political response is yet to be determined. A number of US states are considering imposing resource adequacy or forward contracting obligations that would be placed on the entities that provide retail service to overcome imperfections in wholesale spot markets to restore incentives for investments in generating capacity and demand-response capabilities consistent with traditional reliability levels (California Public Utilities Commission, 2005; Joskow, 2006). The organized markets in the US, Chile, Argentina and Columbia have such obligations or administratively determined capacity payments. These policies are and will continue to attract considerable attention and debate as they should. 9. Environmental policy initiatives can and should be designed to be compatible with competitive wholesale and retail electricity markets. The electric power sector has been a focus of environmental policy for decades. As noted earlier, this has been the case both because of the sectors’ significant impacts on the environment and the political attractiveness of using a regulated industry as a facilitator of environmental policies through “taxation by regulation.” The liberalization of electricity sectors does not appear to have reduced policymakers’ zeal for finding ways to exploit opportunities to exploit taxation by regulation opportunities to pursue environmental agendas. Going forward, environmental policy initiatives affecting the electric power sector need to be compatible with competitive market mechanisms or they will undermine the efficient evolution of competitive markets. Cap and trade programs, as are described in the chapters focused on European countries, for controlling CO2 emissions and as applied in the US for controlling SO2 and NOx emissions (Ellerman et al., 2003) are likely to be compatible with competitive electricity markets if they have the right design features. In particular, as long as emissions permits are fully tradeable, allocations of permits are not updated to reflect changes in emissions, and allowances or credits can be sold when generating plants are retired rather than being recaptured from the plant’s owners by the government, cap and trade programs provide an approach to controlling emissions that is compatible with the development of efficient competitive markets for power. The reliance on tax subsidies, special procurement obligations and contract regimes, subsidized transmission service, and exemptions from balancing and settlement obligations in wholesale markets, all designed to promote renewable energy technologies is of more concern. If we can internalize the relevant externalities with a cap and trade program (or taxes on emissions), the only argument for additional renewable energy subsidies are significant market distortions caused by non-convexities associated with learning by doing economies. However, learning by doing economies are ubiquitous and affect many valuable goods and

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

23

services. Taking this argument to the extreme would imply that the government should subsidize all new products. Moreover, the benefits of subsidies once relevant externalities are internalized should be balanced against the costs of the market distortions that are likely to be caused by the special regime policies governing renewable energy. To the extent that obligations to purchase or pay for above market costs of renewable energy technologies are imposed on retail electricity suppliers, it is important that the obligations apply symmetrically to all electricity suppliers in a symmetrical and non-discriminatory fashion. Otherwise these policies will create an industry devoted to competing for customers by bypassing the costs of renewable energy obligations. 10. Retail market design and the terms and conditions of default service provided by incumbents have important implications for the success of retail competition programs. Consumers can benefit in at least four ways from the introduction of retail competition. First, even if they do not switch to a competitive retailer or competitive electricity service provider (ESP), they may benefit from reductions in regulated “default service”prices that have typically accompanied the restructuring process as an outcome of the bargaining over stranded cost recovery and the terms and conditions under which the incumbents can move their regulated generating plants into unregulated affiliates. Second, consumers can benefit by receiving lower prices than the default service price from an ESP that has competed successfully for their business. Third, ESPs may offer consumers a variety of value-added services, including price risk management, demand management and energy efficiency services. Finally, competing ESPs may be able to provide “retailing” services more efficiently that can the incumbent. However, here we must recognize that retail service costs are a small fraction of a typical customer’s bill, amounting to 0.3–0.4 cents/kWh or about $3–$7 per month for a typical residential customer (depending on assumptions about fixed versus variable components of retail service costs – Joskow, 2000a). Since the incumbent monopolies did not have to incur marketing and advertising costs to attract customers, these are additional costs that are not now reflected in regulated retail prices but would have to be incurred by ESPs. The design of retail competition programs vary widely from country to country and even within countries where reforms have been driven by states and provinces. All countries that have adopted market liberalization reforms allow large customers to buy power competitively at the outset of their restructuring programs. In some countries, retail competition remains available only to such large customers. Residential and small commercial and industrial customers then continue to buy power from their local distribution companies which in turn procure their power in competitive markets and pass along the associated costs in the prices charged to these groups of retail consumers. Other countries have gradually expanded retail competition opportunities to customer classes that consume smaller amounts of power, with the long-run goal of opening up the retail market to all customers. In this case, the distribution company (or a retail affiliate) buys power in the wholesale market and passes along the associated costs to the remaining retail customers during a transition period. Finally, retail competition is sometimes (e.g. in the states in the US that have adopted retail competition programs) made available to all customers at the outset of the reform program. However, since customers, especially smaller customers, do not switch instantly to competitive suppliers, some type of “default service” must be provided to them, typically by their local distribution company or a retail affiliate. Thus, in all cases, there is some period of time during which a significant fraction of retail consumers continue to be served under some type of regulated default service tariff. There are three primary sets of issues associated with the introduction of retail competition that are discussed in the chapters in this volume. First, there are issues associated with the

24

Electricity Market Reform

details of the unbundling and network pricing model applied. Second, there are issues associated with the terms and conditions of retail default power supply services. Third, questions continue to be raised about whether the benefits of making retail competition available to smaller customers is worth the costs relative to the best available wholesale competition alternative. These are all important issues. Unbundling prices and costs for network services from those for competitive electricity supply services, regulating the prices charged for network services and making network services available to all users (customers and retail suppliers) of the network in a nondiscriminatory fashion, are basic and essential steps for retail electricity competition to be successful. Otherwise, the incumbent vertically integrated utility will have an incentive to charge high prices for network services, where it will make a large profit, and low belowmarket prices for power supply services, where it may lose money. This strategy will make it impossible for competitive retail suppliers to compete with the incumbent. This is precisely one of the barriers to retail competition that emerged in Germany in the absence of an electricity sector regulator to establish network access and pricing protocols. To the extent that the incumbent distribution company is also an active unregulated competitor in the retail market, as opposed to a passive supplier of regulated default service that carries no profit margin, competitive retail activities should be fully separated or “ring fenced” from the firm’s regulated activities. The distribution company’s retail affiliate would then look to recover all of its costs, including customer service costs, from revenues received in the market. Functional or ownership separation of retail supply mitigates concerns about cross-subsidization and discriminatory access to information about customers and to the network. This is the approach taken in the UK, Texas and eventually in New Zealand. However, when a distribution company has an obligation to provide “passive” default service, metering, billing and customer services at regulated prices, regulators have an obligation to provide for full recovery of efficiently incurred costs. The treatment of customer service and billing costs is especially challenging since there are significant scale economies in providing these services, making the costs avoided by the incumbent as customers migrate to competitive retailers lower than the average costs of providing these services. As discussed in the chapter by Tschamler on retail competition in the US, the terms and conditions of retail default service can have significant effects on the ability of competitive retailers to attract customers. In the US and some other countries (e.g. Spain), default service prices or tariffs have been used to support a number of objectives other than promoting a robust retail market. These include commitments that retail customers will receive an immediate and sustained price reduction of some magnitude, stranded cost recovery considerations, income redistribution goals and consumer protection goals. As a result, default service prices have sometimes been set at levels below the wholesale cost of power, or wholesale prices have risen over time, closing or reversing the gap between default prices and wholesale market prices. Under these circumstances it is impossible for a competitive retailer profitably to offer services that can attract customers away from default service. Regulatory rules that allow customers to come and go between regulated default service prices and competitive market prices, whichever happens to be lower at a point in time, further undermines the ability of competitive retailers to build a stable customer base. If as a matter of policy regulators want to protect customers from high market prices by giving them access to regulated tariffs fixed at prices below market then retail competition will never be successful. Such policies may also signal a lack of faith and commitment by policymakers in retail competition. The experience in Pennsylvania (a state that is part of the PJM wholesale market) provides a good example of the effects of mixing regulated default pricing with retail competition. Different default service prices were set for each utility in Pennsylvania, reflecting

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

25

Percentage of residential load

historical regulated costs of generation service and stranded cost recovery settlements. The prices were fixed in 2000 for a term of up to 10 years, with some adjustments for fuel and other input price changes. Figure 1 provides time series data on the fraction of residential customers which switched to a competitive retailer for each utility in Pennsylvania. Figure 2 provides the same data for industrial customers. There is both wide variation in the initial fraction of customers who switched to competitive retail suppliers and significant evidence of their switching back and forth between regulated default service and regulated services. The inter-utility variations must be attributable to differences in regulated default service prices since there is no inherent reason why customers in Pittsburgh should be more likely to shop for alternatives than are customers in Philadelphia. By July 2005 nearly all residential customers had returned to regulated default service and a large fraction of the industrial customers who initially opted for default service had also returned to default service. This is attributable to rising nominal wholesale prices in PJM which have reduced or eliminated the “headroom” between the regulated default service price and the wholesale market price for

Pennsylvania direct access load: residential (%)

40

April 2000 July 2000 January 2001 April 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 January 2004 April 2004 July 2004 July 2005

35 30 25 20 15 10 5 0 Allegheny power

Duquesne light

GPU energy

PECO energy

Penn power

PPL

Percentage of industrial load

Fig. 1. Pennsylvania direct access load: residential (%). Source: Pennsylvania office of consumer Advocate.

Pennsylvania direct access load: industrial (%)

80

April 2000 July 2000 January 2001 April 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 January 2004 April 2004 July 2004 July 2005

70 60 50 40 30 20 10 0

Allegheny power

Duquesne light

GPU energy

PECO energy

Penn power

PPL

Fig. 2. Pennsylvania direct access load: industrial (%). Source: Pennsylvania office of consumer Advocate.

26

Electricity Market Reform

power. Accordingly, in the US, the biggest problem faced by competitive retailers is “competition” from default service, a service for which the incumbents typically make no profit either. The general pattern of retail switching behavior in most countries is that large industrial customers are more likely to switch and to do so more quickly than smaller industrial and commercial customers. Residential customers switch more slowly and are more likely to remain with the incumbent, especially when the incumbent must offer a regulated default price that is at or below the wholesale market price of power. However, for residential and small commercial customers, even if the regulated default service price is equal to the comparable competitive wholesale market value of the power supplied, retail suppliers need a significant additional margin both to induce sticky retail customers to switch suppliers and to cover their retail supply costs. This margin has turned out to be much larger than anticipated when retail competition was first introduced. In particular, the retail supply costs for the mass market (residential and small commercial) are much higher than many retailers had anticipated. Billing, customer service, bad debt, advertising and promotion costs add up quickly. Accordingly, the default service price may have to be much higher than the comparable wholesale market price to induce more customer switching. Moreover, the evidence from England and Wales and Texas suggests that price reductions of 5–10% of the total bill compared to the default/incumbent service price are necessary to get significant customer switching for mass market (residential and commercial) customers. If the generation component of the retail price is 50% of the total bill, then price reductions of 10–20% on the generation component are necessary to get significant switching. To this must be added about another 5–10% for retail service costs. So, a margin of 15–30% between the default service price and the comparable wholesale market value of power may be necessary to induce significant switching by residential and small commercial customers. A margin of this magnitude may be incompatible with reducing retail prices from their prevailing or “but for” regulated levels. This naturally leads to the final issue. Is retail competition worth the trouble compared to a regime where the distribution company procures power competitively and resells it at cost to residential and small commercial customers? Unfortunately, there is little if any good empirical analysis available to evaluate this question rigorously, though there is no shortage of strong ideological views. Looking at switching rates alone is not very informative as an index of the welfare consequences of retail competition. The presumption has been that retail competition is a good thing to offer larger customers, where transactions costs are low, opportunities to offer risk management and demand-management products are greater, and customers are expected to be able to shop intelligently. There are also benefits for the development of competitive wholesale market resulting from having more buyers active on the demand side, reducing monopsony problems that might emerge if distributors were the only buyers. Moreover, if the alternative is competitive procurement by the distribution company, regulators must become involved in determining procurement rules, including the attributes of the contracts that will be put out for bids. Industrial customers and their agents should be in a better position to express their risk preferences than are regulators (see Littlechild (2003) for these and other arguments in support of retail competition). And indeed, where default prices have been allowed to float to reflect spot wholesale market prices (including capacity prices), large customers appear to migrate fairly quickly to the market and to sign contracts that hedge price volatility from 1 to 3 years into the future. It is far from obvious to me, however, that residential and small commercial customers have or will benefit much, if at all, from retail competition compared to a regime where their local distribution company purchased power for their needs by putting together a portfolio of short-term forward contracts (from days to several years) acquired in wholesale markets

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

27

(Joskow, (2000a, b) and Littlechild (2003) for a different view)). Indeed, New Jersey has used the so-called basic generation service (BGS) auction process quite effectively to buy power competitively for residential and small commercial customers. There is little evidence that residential and small commercial customers are getting any significant value-added services from retail suppliers aside from some billing options in the UK and, at least in Norway, choices between contracts of different durations. Retail competition with load profiling leads to some inefficiencies (Joskow and Tirole, 2005c). There is evidence that there are significant costs associated with implementing a retail competition program for residential consumers (Green and McDaniel, 1998) and that they may make poor shopping decisions (Salies and Waddams Price, 2004). I remain unconvinced that retail competition for small customers is worth the bother. If policymakers are committed to fostering retail competition for residential and small commercial customers, despite the possibility that retail prices will rise in the short run due to increased transactions costs, switching costs and market power, the framework adopted by the UK, Texas and the Nordic countries is likely to be the most successful in stimulating retail shopping and the development of a viable retail supply sector. 11. Vertical integration between retail supply and generation is likely to be an efficient response to imperfections in wholesale markets. It may also create market power problems. Thus, policymakers must confront a tradeoff: In several countries with active retail competition programs there appears to be a growing movement to an industry structure where competitive retail suppliers acquire generating capacity to meet a significant fraction of their retail commitments. This trend is likely to reflect an efficient response to relatively high transaction costs associated with real wholesale power markets in practice (Coase, 1937; Williamson, 1975; Carlton, 1979). There is no inherent competition problem with vertical integration of this type as long as there are a sufficient number of vertically integrated suppliers that continue to compete in the market. However, if there is significant market power in the upstream or downstream markets, vertical integration could lead to a further reduction in competition by increasing the operating or entry costs of rival retail suppliers (Ordover et al., 1990; Riordan, 1998). The discussion of retail competition in the chapter on New Zealand suggests that the intensity of competition declined significantly as retail suppliers became vertically integrated while the chapter on Australia suggests that vertical integration did not lead to market power problems there. Thus, there may be a tradeoff between increases in efficiency and increases in market power. The welfare properties of retail competition with different horizontal and vertical market structures has received too little serious theoretical and empirical analysis and more work on these issues would be desirable. 12. Expanding demand response in spot wholesale energy markets needs more attention. In markets for most goods and services, when demand grows and supply capacity constraints are reached, prices rise to ration demand to match the capacity available to provide supplies to the market. In electricity markets, however, as generating capacity constraints are reached, relatively little demand can be rationing by short-term price movements and, instead, must by rationed with rolling blackouts. This reflects both the limited use of real time pricing and the system operator’s need to adjust demand very quickly at specific locations. The possibility of broader uncontrolled cascading blackouts and regional network collapses further exacerbates this problem and necessarily leads to regulatory requirements specifying operating reserves, operating reserve deficiency criteria and associated administrative actions by system operators to balance the system to meet voltage, stability and frequency requirements in an effort to avoid cascading blackouts (Joskow and Tirole, 2005b). In addition, retail competition has

28

Electricity Market Reform

more attractive welfare properties if the real time consumption of retail consumers can be measured instead of relying on load profiling (Joskow and Tirole, 2005c). The challenges faced by network operators to maintain system reliability and avoid nonprice rationing of demand would be reduced if additional demand-side response instruments were at their disposal. These instruments include the ability to rely on demand response by more customers who can see and respond to rapid changes in market prices and expanded use of price-contingent priority rationing contracts (Chao and Wilson, 1987). As a general matter, too little demand-side response has been developed to date most countries. (Texas has quite a bit of demand response but this is likely to be associated with the relatively large number of oil refineries, petrochemical and other energy intensive industries that can manage their load and have self-generation facilities that can be used when market price rise sufficiently.) The demand-response instruments that are available are poorly integrated with spot markets and are likely to have the effect of depressing prices inefficiently. Moreover, the prices that are paid for demand response or the prices that can be avoided by responding to price signals are too low compared to the cost of carrying generating capacity reserves to meet planning reserve margins in some cases. Improving demand response should be given higher priority in wholesale market design (Borenstein, Jaske and Rosenfeld, 2002). There is no reason why these efforts must be limited to large customers. Controlled air conditioning and water heater devices have been available to residential and small commercial customers in several countries for many years. Italy’s decision to install real time meters for every electricity consumer premises is an important step forward in this area. By making real time metering universal, Italy will gain the benefits of an infrastructure that can support additional efficient demand response and avoid the impediments to retail competition identified in the chapter on the Nordic markets. 13. Electricity sector reform appears to be a continuing process of improvement, but a process of continuing reforms of the reforms has both potential benefits and potential costs. It is quite clear from the chapters in this book that none of the reform programs got it all right out of the box. Initial reform programs are followed by additional reforms, some major and some minor, to respond to performance problems that emerge in practice or lessons learned about best practices from other countries. On the one hand, reforms that are needed to fix major performance problems certainly should be considered carefully. On the other hand, a process of ongoing reforms that have significant and uncertain future financial impacts on market participants is not likely to create a framework that is conducive to investment in long-lived assets whose value is subject to policy reform risks. Policy reforms may also be used opportunistically to respond to political pressures that arise under market conditions when investors properly expected that they would achieve high returns, effectively truncating the upper end of the return distribution and leading investors to require higher expected returns from other states of nature than would otherwise be the case. So, for example, it would be unfortunate if policymakers were to require renegotiation of restructuring arrangements because natural gas prices have risen dramatically, benefiting deregulated coal, nuclear and hydroelectric facilities. Similarly, if policymakers belatedly conclude that there are market power problems, the tools available to them are likely to be much more limited after assets have been privatized, deregulated, traded, constructed, etc. than before, if they are to limit themselves to policies that do not amount to expropriation of private property rights that they previously conveyed. Policy reform risk is extremely difficult to hedge, except perhaps through the regulatory process itself and is a potentially significant deterrent to investment. Accordingly, it makes a lot of sense to try to get the reform program as correct as possible the first time around. It is also important for policymakers to recognize that the search for

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

29

perfection can be the enemy of the good. Policymakers need to make sure that the benefits of any additional reforms exceed their short- and long-run costs, in particular those related to investment incentives. And if there are to be reforms of the reforms it is desirable to package them together so that there can be one reform of the reforms rather than a continuing stream of them. Finally, if policymakers are serious about competitive markets for power they will have to rethink the long tradition of relying on taxation by regulation of the electric power industry to implement policies in ways that hide the associated costs from taxpayers. 14. A strong political commitment to reform is important. Implementing a good electricity sector liberalization program is a technical, institutional and political challenge. Almost everywhere, some unanticipated (at least by the policymakers) problems emerged that required major or minor refinements to the original reform program. In some cases (e.g. UK, New Zealand, Alberta, Australia, Texas) the reforms were consistent with the continuing development of competitive markets and in other cases they were not (e.g. California, Ontario, Brazil). It appears that reforms that have strong pro-competition political support are more likely to respond to problems by identifying market or institutional imperfections and trying to fix them in ways that are consistent with the continued successful evolution of competitive wholesale and retail markets. They are also likely to be willing to live with some imperfections, recognizing that no market is perfect and that the cures can be worse than the disease. Where the commitment to competitive electricity markets is weak, when problems emerge policymakers are more likely to seek what appear to be quick fixes that undermine continued evolution of competitive markets or just cut and run from the competitive market agenda. If the commitment to competition is not strong in the first place, of course, the reforms are likely to be timid and have little effect on the status quo anyway, Japan and many US states being the prime examples. If policymakers do not have a strong commitment to competition and are unable to follow the basic textbook reform model outlined at the beginning of this chapter, sector performance may be better if they focus their attention on improving the traditional regulated monopoly model rather than dabbling with timid and flawed approaches to competition.

Conclusion Structural, regulatory and market reforms have been applied to electricity sectors in many countries around the world. Significant performance improvements have been observed in some of these countries as a result of these reforms, especially in countries where the performance of state-owned monopolies was especially poor. Privatization and PBR mechanism applied to regulated distribution companies has generally yielded significant cost reductions without reducing service quality. Wholesale markets have also stimulated improved performance from existing generators and helped to mobilize significant investments in new generating capacity in several countries. However, efforts to create well-functioning competitive wholesale and retail markets have revealed many significant challenges and the restructuring and competition reforms remain a work in progress in most countries. The California electricity crisis, electricity crises in Brazil, Chile, Ontario, and elsewhere, scandals involving energy trading companies like Enron, the failure of poorly designed reforms in countries such as Brazil, macroeconomic problems undermining investments in generally well-designed systems as in Argentina, and ongoing political interference undermining private sector investments as in India and Pakistan, have certainly made policymakers more cautious (but not necessarily more thoughtful) about

30

Electricity Market Reform

electricity sector reforms. The challenges associated with successful reforms have sometimes been underestimated. However, these problems and challenges do not imply that restructuring, regulatory reform, and promoting the development of competitive wholesale and retail markets for power, are ill-advised. The problems that have emerged are now much better understood and solutions to many of them are at hand. The primary question is whether governments properly can choose between competing solutions and have the political will to resist interest group pressures and pursue reforms that will lead to more efficient markets and better performance of the network platforms upon which competition depends.

References Bacon, R.W. and Besant-Jones, J.E. (2001). Global electric power reform, privatization and liberalization of the electric power industry in developing countries. Annual Reviews of Energy and the Environment, 26, 331–359. Beesley, M. and Littlechild, S. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20(3), 454–472. Bertram, G. and Twaddle, D. (2005). Price–cost margins and profit rates in New Zealand electricity distribution networks: the cost of light handed regulation. Journal of Regulatory Economics, 27(3), 281–307. Besant-Jones, J.E. (ed.) (1993). Reforming the Policies for Electric Power in Developing Countries. Washington, DC: World Bank. Borenstein, S., Bushnell, J. and Wolak, F. (2002). Measuring market inefficiencies in California’s restructured wholesale electricity market. American Economic Review, 92(5), 1376–1405. Borenstein, S., Jaske, M. and Rosenfeld, A. (2002). “Dynamic Pricing, Advanced Metering and Demand Response in Electricity Markets,” Center for the Study of Energy Markets Working Paper, University of California at Berkeley, October. California Public Utilities Commission (2005). “Capacity Markets White Paper,” August 25. Carlton, D. (1979). Vertical integration in competitive markets under uncertainty. Journal of Industrial Economics, 27, 189–209. Chao, H. and Wilson, R. (1987). Priority service: pricing, investment and market organization. American Economic Review, (77), 899–916. Coase, R. (1937). The nature of the firm. Economica, 4, 386–405. De Araujo, J.L.R.H. (2001). Investment in the Brazilian ESI – What Went Wrong? What Should Be Done? Institute of Economics, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil. Domah, P.D. and Pollitt, M.G. (2001). The restructuring and privatisation of the regional electricity companies in England and Wales: a social cost benefit analysis. Fiscal Studies, 22(1), 107–146. Ellerman, A.D., Joskow, P.L. and Harrison, D. (2003). Emissions Trading in the United States. Pew Center on Global Climate Change, Washington, DC. Estache, A. and Rodriguez-Pardina, M. (1998). “Light and Lightening at the End of the Public Tunnel: The Reform of the Electricity Sector in the Southern Cone,” World Bank Working Paper, May. Gilbert, R., Neuhoff, K. and Newbery, D. (2002). “Allocating Transmission to Mitigate Market Power in Electricity Networks, Cambridge-MIT Electricity Project Working Paper, October. Green, R. and McDaniel, T. (1998). Competition in electricity supply: will “1998” be worth it? Fiscal Studies, Vol. 19, No. 3, pp. 273–293. Green R. and Newbery, D. (1992). Competition in the British Electricity Spot Market. Journal of Political Economy, 100(5), 929–953. Hogan, W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4, 211–242. Hunt, S. (2002). Making Competition Work in Electricity. New York, Wiley. ISO New England (2005). “2004 Annual Markets Report.” http://www.iso-ne.com/markets/ mkt_anlys_rpts/annl_mkt_rpts/2004/2004_annual_markets_ report.doc Jamasb, T. (2002). Reform and Regulation of the electricity sectors in developing countries, June 2002, Department of Applied Economics, University of Cambridge.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

31

Joskow, P.L. (1989). Regulatory failure, regulatory reform and structural change in the electric power industry. Brookings Papers on Economic Activity: Microeconomics, 125–199. Joskow, P.L. (1997). Restructuring, competition and regulatory reform in the US electricity sector. Journal of Economic Perspectives, 11(3), 119–138. Joskow, P.L. (1998). Electricity sectors in transition. Energy Journal, 19, 25–62. Joskow, P.L. (1998a). Electricity sectors in transition. Energy Journal, 19, 25–62. Joskow, P.L. (2000a). Deregulation and regulatory reform in the US electric power sector. In Sam Peltzman and Clifford Winston (eds.), Deregulation of Network Industries: What’s Next? Brookings Institution Press, Washington, DC. Joskow, P.J. (2000b). Regulating the electricity sector in Latin America, comments 2000b. Economia, 1, 199–215. Joskow, P.L. (2001). California’s electricity crisis. Oxford Review of Economic Policy, 17(3), 365–388. Joskow, P.L. (2003). Electricity sector restructuring and competition: lessons learned. Cuadernosde Economia (Latin American Journal of Economics), 40, 548–558. Joskow, P.L. (2005a). The difficult transition to competitive electricity markets in the United States. In J. Griffin and S. Puller (eds.), Electricity Deregulation: Where To From Here? University of Chicago Press, Chicago. Joskow, P.L. (2005b). Transmission policy in the United States. Utilities Policy, 13, 95–115. Joskow, P.L. (2005c). Regulation and deregulation after 25 years: lessons for research. Review of Industrial Organization, 26, 169–193. Joskow, P.L. (2005d). “Incentive Regulation in Theory and Practice: Electric Transmission and Distribution,” NBER Regulation Project, August. Joskow, P.L. (2006). Markets for power in the US: an interim assessment. The Energy Journal, 27(1), 1–36. Joskow, P.L. and Kahn, E. (2002). A quantitative analysis of pricing behavior in California’s wholesale electricity market during summer 2000. The Energy Journal, 23(4), 1–35. Joskow, P.L. and Rose, N.L. (1989). The effects of economic regulation. In R. Schmalensee and R. Willig (eds.), Handbook of Industrial Organization. North-Holland, Amsterdam. Joskow, P.L. and Schmalensee, R. (1983). Markets for Power: An Analysis of Electric Utility Deregulation. MIT Press, Cambridge. Joskow, P.L. and Tirole, J. (2000). Transmission rights and market power on electric power networks. Rand Journal of Economics, 31(3), 450–487. Joskow, P.L. and Tirole, J. (2005a). Merchant transmission investment. Journal of Industrial Economics, 53(2), 233–264. Joskow, P.L. and Tirole, J. (2005b). Reliability and competitive electricity markets, September (revised) http://econ-www.mit.edu/faculty/download_pdf.php?id⫽917. Joskow, P.L. and Tirole, J. (2005c). Retail electricity competition, Rand Journal of Economics (forthcoming) http://econ-www.mit.edu/faculty/download_pdf.php?id⫽918. Littlechild, S.C. (2003). Wholesale spot market passthrough. Journal of Regulatory Economics, 23(1), 61–91. Newbery, D. and Pollitt, M. (1997). The restructuring and privatization of Britain’s CEGB – was it worth it? Journal of Industrial Economics, 45(3), 269–303. Ordover, J., Salop, S. and Saloner, G. (1990). Equilibrium vertical foreclosure. American Economic Review, 80, 127–142. Pollitt, M. (2004). “Electricity reforms in Argentina: lessons for developing countries. CMI Working Paper 52, Cambridge Working Papers in Economics. http://www.econ.cam.ac.uk/electricity/publications/ wp/ep52.pdf Riordan, M. (1998). Anticompetitive vertical integration by a dominant firm. American Economic Review, 88, 1232–1248. Rudnick, H. (1996). Pioneering electricity reform in South America. IEEE Spectrum, August, 38–45. Rudnick, H. (1998). Market restructuring in South America. IEEE Power Engineering Review, June, 3–6. Rudnick, H. and Zolezzi, J. (2001). Electric sector deregulation and restructuring in Latin America: lessons to be learnt and possible ways forward. IEEE Proceedings Generation, Transmission and Distribution 148, 180–184.

32

Electricity Market Reform

Salies, E. and Waddams Price, C. (2004). Charges, costs, and market power: the deregulated UK electricity retail market. The Energy Journal, 25(3), 19–37. Schmalensee, R. (1989). Good regulatory regimes. Rand Journal of Economics, 20(3), 417–436. Stoft, S. (2002). Power System Economics. Wiley Interscience, New York. Williamson, O. (1975). Markets and Hierarchies: Analysis and Antitrust Implications. Free Press, New York. Williamson, O. (1985). The Economic Institutions of Capitalism. Free Press, New York. Wolfram, C. (1999). Measuring duopoly power in the British electricity spot market. American Economic Review, 89(4), 805–826. World Bank (1994). Infrastructure for Development: World Development Report 1994.

PART I What’s Wrong with the Status Quo?

This page intentionally left blank

Chapter 1 Why Restructure Electricity Markets? FEREIDOON P. SIOSHANSI1 AND WOLFGANG PFAFFENBERGER2 1

Menlo Energy Economics, Walnut Creek, CA, USA; 2International University Bremen, Bremen, Germany

Summary For nearly two decades, governments in a number of countries have restructured and/or liberalized their electricity supply industry. The motivations to change the regulatory regime and the expectations for the outcomes vary despite many commonalities. The experience of different countries, however, has been decidedly mixed, highly to mildly successful in some places, disappointing or disastrous in others. Understanding why some restructured markets are functioning successfully while others have not is the central theme of this book, which offers a variety of perspectives from different parts of the world. This chapter provides a review of the shortcomings of the traditional regulated monopoly structure, examines the motivations for restructuring and the expectations of policy-makers in introducing market reforms.

1.1. Why Change the Status Quo? As the following chapters of this book point out, before the restructuring developments of 1980s and 1990s, the electricity supply industry (ESI) was generally organized and operated under one or a combination of the following two structures: ● ●

as state-owned enterprises (SOEs), or as privately owned, regulated monopolies.

In the former case, typically there was no independent regulator but a separate part of the government (e.g. the treasury) might exercise regulatory oversight or provide financial or fiduciary function. This model has some shortcomings. The taxpayers, for example, usually bear most investment risks and there may be poor accountability since the government agency is not directly accountable to consumers or shareholders. Moreover, under this scenario there tends to be some circularity, since different arms of government would be engaged in forecasting, planning, building, investing and operating as well as managing the network, setting and collecting retail tariffs. SOEs may not be overly sensitive to customer needs and may 35

36

Electricity Market Reform

lack sufficient incentives to improve customer service or engage in technology innovation. In the case of many rapidly growing economies, central government may not have sufficient resources to adequately invest in infrastructure, resulting in chronic power shortages and poor service reliability. In the case of privately owned, regulated monopolies, which are prevalent in more advanced economies, the private sector owns and operates some or significant components of the ESI under the supervision of an independent regulator. Regulated monopolies could potentially offer some advantages relative to the centrally planned SOEs especially when there is a competent, resourceful, independent and vigilant regulator. At least in theory, this model could capture the significant economies of scale of large vertically integrated monopolies while controlling their abusive tendencies. Regulated monopoly regime could provide a degree of price stability and encourages long-term planning. In some countries, this model has resulted in the formation of a number of large vertically integrated companies, which – according to some studies – offer significant cost savings.1 America, Germany and Japan, which are covered in the following chapters, offer examples of this regulatory paradigm, with a small number of large vertically integrated utilities dominating the business. Despite these purported advantages, the regulated monopoly model has a number of shortcomings, including the following: ●

●

●

●

●

1

Over-investment in rate base: Under traditional rate of return regulations in the USA, firms are rewarded for how much they have invested in infrastructure. Since the rate of return is typically fixed, regulated monopolies, all else being equal, prefer to have more assets in the so-called “rate base.” This leads to a perverse tendency to over-invest, if permitted by the regulator, who typically has to approve capital investments in advance.2 Security of supply: In the regulated monopoly model, the trade-off between quality and price was usually biased toward security of supply as customers had no choice. As will be described in several chapters of this volume, in the transformation process to a marketoriented model, this may cause some problems as customers, who are used to taking security of supply for granted, may not be conscious about the cost of quality or reliability.3 Risks borne by ratepayers: Regulated monopolies typically pass all investment risk to captive customers, who must bear it during the long life of capital assets. No customer choice: In exchange for “the obligation to serve”, regulated utilities get an exclusive service area, which means that customers have no choice in selecting a service provider even if a lower-cost option is available.4 While this may not be much of an issue for most small consumers, it may be a source of frustration for large, sophisticated consumers. Price disparities: In many countries where neighboring areas are served by different utilities and/or under different regulatory regimes, significant price disparities have evolved

For a comprehensive discussion of the benefits of vertical integration see Michaels (2005) and his references. 2 For example, see the seminal paper of Averch and Johnson (1962). 3 Kollmann, Andrea; Tichler, Robert; Schneider, Friedrich: Netztarife in Österreich: Bestandsaufnahme und Interantionaler Vergleich, Linz (2005) show that there is a strong correlation between network fees and quality of service. 4 Competition also forces unbundling of costs of service components and the removal of cross-subsidies among different customer classes – a desirable development from efficiency point of view. It also encourages suppliers to pay closer attention to customer demand and offer services tailored to their changing needs.

Why Restructure Electricity Markets?

●

37

over time, causing significant pressure for change. Yet rigid regulations typically do not allow customers to switch suppliers. Price subsidies: In many countries, the regulators, usually with tacit approval of politicians, find it convenient to allow, or encourage, cross-subsidization of rates between different customer classes. The same applies to various cost components of bundled service, making it difficult, for example, to decipher the cost of energy versus transmission services, versus distribution. As will be pointed out, competition encourages and/or requires costs to be unbundled and be divulged, making it difficult to cross-subsidize certain customer classes or provide cross-subsidies between various service components.

On the other hand, the regulated model has a number of other attributes with potentially significant consequences for efficiency and price transparency: ●

●

●

5

Supply adequacy: Both SOEs and regulated monopolies are subject to a highly skewed reward and penalty system, which predisposes them to over-build. The reason is obvious. There will be a public inquiry and an outcry if there is a shortage of capacity, while hardly anyone notices if there is a little extra capacity sitting idle. One is tempted to conclude that because of this phenomenon, regulators rarely had to intervene in the past to ensure adequate reserve capacity – the ESI would always err on the side of having a little too much. By contrast, the issue of who is responsible for resource adequacy has become a hot topic in a number of countries since the introduction of competition.5 Manipulation by politicians: SOEs and regulated monopolies offer irresistible opportunities for regulators and politicians to micro-manage and/or “meddle” in the business.6 Examples include efforts to manipulate prices, offer subsidies to certain constituents, or engage in other political or social experiments – which would be hard to justify in a competitive environment with publicly listed companies. For example, in many Indian provinces, farmers pay next to nothing for electricity – as a matter of political expediency. In Britain before liberalization, the coal unions enjoyed strong political influence, which included the use of uneconomic British coal by government-run Central Electricity Generating Board (CEGB).7 In many US jurisdictions, commercial and industrial customers routinely subsidize residential customers. In California, for example, residential customers in hot interior climate zones are subsidized by those living in the more temperate coastal zones. Similar examples abound in other countries. It must be noted that governments continue to meddle in the utility business even after the firms are privatized and operate competitively – by imposing, for example, renewable portfolio standards, energy efficiency mandates and other measures. Nuclear energy: Capital-intensive nuclear industry owes special thanks to the relative long-term stability of the regulated environment. Without long-term contracts and/or

In the USA, for example, the Federal Energy Regulatory Commission (FERC), devoted a major section to discussion of supply adequacy in its standard market design (SMD) proposal in 2002. The issue of supply adequacy is covered in a number of chapters in this book. 6 An example of political meddling, not unique by any measure, occurred in late 2003 in Italy, where the independent regulator, the Authority for Electricity and Gas, proposed to reduce ENEL’s average rates by 2% over a 4-year period just as the government was trying to partially privatize the company. The regulatory decision had a negative effect on ENEL’s public offering, then 61% owned by the government. The government subsequently put intense pressure on the regulator to “rethink” its earlier decision, making a mockery of its independence. 7 See chapter on Britain in this volume for further details.

38

●

Electricity Market Reform government assurance, private investors would shun large, lumpy investments with significant upfront investments, a hallmark of nuclear power plants.8 Public disclosure and coordination: Another potential advantage of regulation, not fully appreciated until recently, is that it may encourage public disclosure as a part of the regulatory process while enforcing some coordination in forecasting and planning.9

Under a “deregulated” paradigm, where vertically integrated utilities are typically broken apart, there may be little coordinated planning. The problem, some believe, is particularly acute due to the bifurcation of generation and transmission planning – which now takes place within different, often competing, companies.10 Moreover, in a competitive environment, there may be strong disincentives in communication and coordination.11 In the North American context, recent “boom and bust” cycles in capacity investment may have been exacerbated by the introduction of competition in the wholesale market. In the recent past, there have been significant capacity gluts in certain regions and subsequent drops in wholesale prices. Wholesale competition has also been blamed for increased transmission congestion problems since transmission and generation planning have become virtually disjointed. In some cases, the result has been significant local capacity build-ups while shortages afflict other areas. In some regions, long-term resource planning has become chaotic with concerns about who is responsible for maintaining adequate reserve margins, as a number of chapters in this book point out.12 Further research appears warranted to identify and quantify the existence and prevalence of these problems and whether they impose significant risks. As the chapter on Europe claims, the introduction of competition on the basis of the 1996 European Commission directive may have resulted in dampened interest in generation investment. This, in turn, has raised the significance of “security of supply,” which is further described in several chapters, notably in case of Germany. In the European context, current projections show potential problems on the horizon regarding supply adequacy (Fig. 1.1). It will be a test for the market to see if additional generation capacity will develop in the new framework.

8 The chapter on Japan describes the difficulties facing Japanese regulators in defining a sustainable future for nuclear power, which has significant energy security implications for Japan. 9 One does not need a regulator for this. For example, the California Energy Commission maintains public records on all plants under construction in California regardless of their ownership including IPPs and co-generation units that are not regulated. 10 There are those who point out that coordination is not all that complicated, and future and forward contracts can substitute for lack of ownership of the long supply chain. This is true in theory but not always in practice, for example, in cases where generators cannot secure long-term contracts for their output, or secure transmission rights to transmit their output to distant markets. In the case of wind power, for example, developers typically select wind parks with little concern about availability of transmission capacity or distance to load centers. 11 In fact, it can be argued that competing generators have every incentive to confuse their potential rivals by spreading false rumors about their future investment plans, bluffing or making misleading announcements designed to confuse and distract. The last thing a private generator wants to do is to spell out its plans to its rivals. 12 Under regulation, non-competing utilities had strong incentives to coordinate their plans and make sure that the “neighborhood” had sufficient capacity to maintain reliable service and to keep the regulators at bay.

39

Why Restructure Electricity Markets? GW 70,0 65,0

62,0

62,4

60,5

(60,5)

(60,6)

60,0

(56,3) 55,0

55,0 50,0

50,2

51,6

53,0

(53,4)

53,8

(54,4)

January 2008

January 2009

52,0

45,0 January 40,0 35,0 30,0 January 2004

January 2005

January 2006

Remaining capacity

January 2007

January 2010

5% of GC ⫹ margin against peak load

Fig. 1.1. Projections of installed capacity and peak demand for Europe, January 2004–2010, in GW. Source: UCTE.

1.2. Drivers of change in regulatory paradigm Starting in 1980s, some economists and policy-makers started to question the wisdom and the necessity of centrally planned SOEs and regulated monopolies. A number of factors contributed to the new thinking13 including one or more of the following – not necessarily in any order and not applicable in all cases: ●

●

●

13

Gas turbines: The advent of highly efficient natural gas-fired turbines made it possible to build smallish units in record time with little risk. This broke significant barriers to entry in generation and made large, capital-intensive plants less attractive. As documented in Figure 1.2 for the USA, the dominance of gas-fired technology has been overwhelming in the recent past, and the phenomenon is not unique to America. Sympathetic regulators in a number of countries have removed the barriers for newcomers to enter the generation business and compete with the incumbent regulated utilities. Independent power producers (IPPs) have made significant inroads in many markets, in many cases, using highly efficient gas turbines. Ideology and politics are believed to have played a role in some cases.14 Privately listed companies, under commercial competitive pressures, can be expected to trim bloated staff levels, shed uneconomic contracts, and be forced to divulge performance information.

Among 62 countries studied, Pollitt (1997) found liberalization in 51, privatization in 30 and vertical separation in 27. The interest in market reform, by all indications, has intensified since 1997. 14 Newbery (1999) says, “one rather cynical view of electricity privatization (in Britain) is that it was part of a campaign against the overmighty public sector unions which was designed to undermine the power base of the opposition labor party.” The number of British coal miners which stood around 160,000 in 1984 dropped to 10,000 in 1994, and has continued to fall.

40

Electricity Market Reform 200

186

180 160 140 GW

120 100 80 60 34

40

11

20 0

Natural gas

Coal

Wind

6 Hydro

15 Other

Fig. 1.2. The dominance of gas-fired turbines is evident in the US context. New capacity built or proposed in the US between 2003 and 2007 in GW. Source: Platt’s RDI.

●

●

●

● ●

15

Public debt may have been a motivating factor in a few cases, such as in Victoria, Australia, where the sale of state-owned assets brought relief to the heavily indebted government of the time. Regulatory complexity played a role in California, where state regulators reached the humbling conclusion that, despite their best efforts, California electricity prices were 50% above national average. In this case, the regulators conceded that they could no longer do an adequate job of regulating the utilities, leading to the decision to “restructure” the industry to allow market discipline to regulate the industry instead.15 As further described in Chapter 10, the restructuring did not work very well, resulting in even higher prices for California consumers. Inadequate investment in infrastructure is among the primary reasons for many rapidly growing economies, which privatize and/or liberalize their ESI to attract foreign investment.16 Poor accountability by SOEs is among the motivating factors in some cases.17 Decentralized decision-making is a motivator in cases where central government can no longer cope with the growing complexities of forecasting, financing, constructing, operating and maintaining the network.18

In mid-1990s, the IOUs in California where encumbered by heavy debt having completed a number of costly nuclear power plants while the new IPPs were building highly efficient natural gas-fired gas turbines at a time when natural gas prices were low. This led to protracted discussion about “stranded costs,” which, in retrospect proved to be illusory. Refer to Chapter 10 for further details. 16 Financial lending institutions including The World Bank, the Inter-American Development Bank (IADB) and The Asian Development Bank (ADB) were unanimous in recommending privatization as a way to attract foreign investment in infrastructure, sometimes without adequate regard to the local underlying institutional constraints leading to significant problems in implementation. 17 In some countries like France and Korea, the dominant position of the monopoly utility company has simply been overwhelming. 18 In Iran, a frustrated official related a story that the head of a provincial generating plant had called to inform him that the plant would be shut down in a few days time because the fuel in the storage tanks

Why Restructure Electricity Markets?

41

Box 1.1 Terminology Terminology used to describe different approaches to change the regulatory paradigm: Restructuring is a broad term, referring to attempts to reorganize the roles of the market players, the regulator and/or redefine the rules of the game, but not necessarily “deregulate” the market. Liberalization is not synonymous with restructuring. It refers to attempts to introduce competition in some or all segments of the market, and remove barriers to trade and exchange. The European Union, for example, refers to their efforts under this umbrella term. Privatization generally refers to selling government-owned assets to the private sector, as was done in most countries that have embarked on market reform. Corporatization generally refers to attempts to make SOEs look, act and behave as if they were for-profit, private entities. In this case, a SOE is made into a corporation with the government treasury as the single shareholder. For example, former SOEs in New South Wales, Australia, have been corporatized. They vigorously compete with one another, while all belong to the same, single shareholder, namely the Government Treasury of NSW. Deregulation refers to removing or reducing sector-specific regulation and subjecting the ESI to the monitoring by the anti-cartel authority. However, no electricity market has been (or, in fact, can be) fully deregulated. There is agreement now that the monopolistic bottlenecks of the market in transmission and distribution need specific regulation in addition to general anti-cartel policies.

Depending on the real or perceived problems with the status quo, and what was viewed as the solution, a different approach was pursued in different countries.19 The following chapters of this book offer a number of examples from different countries on what prompted a change in the regulatory paradigm, how the new market structure was designed and implemented, and what have been the results thus far. Box 1.1 provides definitions of a few key words. 1.3. Alternative views on competition and regulation In most cases, the shift in regulatory paradigm includes a desire to encourage competitive market forces to substitute for command and control regulations, or bureaucratic and often inefficient management of SOEs. Since some segments of the ESI are natural monopolies and cannot be made competitive, the focus of market reform has been predominantly on generation and supply functions, which can be competitively provided.20 The question is how to make the whole system work efficiently, while parts of it remain regulated. 18

(continued) was running short, pending shipments ordered by the central authorities. The manager could not imagine taking the matter into his own hands and securing additional fuel. Examples like this abound in countless other centrally planned economies. 19 Jamasb and Pollitt (2005) provide a comprehensive discussion of the European market reform initiatives. 20 Refer to Chapter 2 as well as Jamasb and Pollitt (2005) for a discussion of these issues.

42

Electricity Market Reform

This book’s Foreword, Introduction and Chapter 2 offer useful insights gained after two decades of studying markets. The following chapters provide a variety of models and market structures. At a fundamental level, there may be two basic views on competition and the role of regulation in the ESI: ●

●

Create level playing field where the regulator plays the role of the referee: Those who subscribe to this view see the prime purpose of restructuring as devising a set of rules and conditions that encourage free and fair competition among the players – and believes that the main role of the regulator is to be an impartial referee. This implies regulating only the monopolistic bottlenecks of the industry and deliberately refraining from regulating generation or supply. There are good reasons for such a strategy as Knieps points out in Chapter 2. A number of authors in this book favor a light-handed regulatory philosophy, leaving as much to markets as markets can reasonably handle. Littlechild, for one, believes in “competition where possible, regulation where not.”21 Write a script and make the puppets dance: Those who subscribe to this view prefer a more prescriptive market and a more hands-on role for the regulator. In his case, the regulator prepares a prescriptive plot for the play, making sure that the players act according to the script. As noted by Knieps in Chapter 2, “It is well known from the positive theory of regulation that regulators have strong incentives to over-regulate, mix regulatory instruments in an unsuitable way, favor the application of detailed regulation and call for a heavy-handed supervision of firms.” Authors in some chapters appear to favor the strong hand of regulator over the invisible hand of the market.

The central ingredients of the former are well-known, namely: ● ● ● ●

● ●

●

Remove barriers to entry in generation.22 Privatize or corporatize all players so they are competing on the same footing.23 Unbundle vertically integrated enterprises to remove cross-subsidies and self-dealing.24 Ensure that the transmission network is open and accessible to all under transparent and non-discriminatory prices.25 Create wholesale markets that are open and transparent. Ensure that the grid is managed by an independent operator who maintains reliability, manages transmission congestion, operates various markets to facilitate trade, liquidity and risk management. Foster competition in the supply business.

These conditions will most likely encourage a number of firms to enter various segments of the business, engage in short- and long-term transactions with one another and with customers. With sufficient degrees of freedom, the participants will find suitable arrangements to transact, and the regulator’s role is primarily that of a vigilant referee, making sure that 21

See Littlechild (2005). In the US context, the Public Utility Regulatory Act (PURPA) of 1978 accomplished this goal to a great extent. 23 In the Australian context, Victoria has done this while the neighboring New South Wales has maintained government control of both generation and distribution, creating an uneven playing field especially in the retail sector. 24 As pointed out in Chapters 9 and 18, lack of rigorous physical unbundling is considered as a major obstacle the creation of a vibrant competitive market. 25 Chapters 9 and 14 address the problems associated with this thorny issue in the European and US context. 22

Why Restructure Electricity Markets?

43

the rules of the game are obeyed, infringements are caught and offenders are punished. Otherwise, market participants are free to roam as long as they obey by the rules and play within the field. Most restructured markets have adopted a variation of this theme – with varying levels of autonomy and authority for the referee. This assumes that the players know best, and the market is simply the sum of its components. Free market advocates and those favoring laissez faire prefer a limited role for the regulator, allowing market participants maximum flexibility.26 The alternative view of the market assumes a stronger, perhaps intrusive role for the regulator, most likely requiring frequent intervention in the market. In this case, the roles of the players are carefully orchestrated and their moves are monitored and controlled. There is, of course, no assurance that the scriptwriter, no matter how clever, can get all the parts right. Nevertheless, examples of this line of thinking persists and may be found in some countries, notably those who have had a disappointing experience with market liberalization and reform. There may also be considerable public pressure on governments to replace the invisible hand of the market by the intrusive hands of government. The degree of government interference is part of the political culture of a country that has evolved over a long time and often is amazingly stable. Tradition, to some extent, explains the deviation between the textbook approach to restructuring and the implementation of reforms in different countries.

1.4. In search of insights The following chapters of this book expand on many of the topics identified here while exploring others. Many important answers are provided in the context of examining specific experiences of particular markets around the world. While many insights are offered, a number of questions remain essentially unanswered, either because we have not examined the issues or have not found the answers. Additionally, three important caveats must be mentioned at the outset: ●

●

●

First, the co-editors of this volume made a conscious decision not to influence the personal perspectives of individual authors or to restrict their views in any way. Consequently, this volume offers a wide range of philosophies on markets, which the reader should find refreshing. Second, while we have made an effort to cover most interesting restructured markets, not all are included. Third, we have purposely limited the scope of the discussions to market design, structure and performance issues – leaving out other important topics such as renewable energy technologies, global climate change, environmental and energy security issues, to name a few.

In this book’s Foreword and Introduction, Stephen Littlechild and Paul Joskow share their own considerable experiences and perspectives, as well as an overview of many market design issues covered in this book. In Chapter 2, Guenter Knieps presents the analytical foundation of network-specific market power and provides a survey of the localization of monopolistic bottlenecks in different networks. Modern economic theory has developed the concept of disaggregated regulation

26

For example, generators would not be obliged to bid their capacity in the market, nor would they be required to explain their bidding strategies or prices.

44

Electricity Market Reform

of monopolistic bottlenecks against end-to-end regulation. He shows that there are solid reasons that regulation should be restricted to the bottleneck part of the network making sure that markets for complementary services like generation and supply will be able to breathe. As far as market power in these fields is concerned, it should be subjected to general competition policy rather than by sector-specific regulation. In Chapter 3, Ricardo Raineri describes the pioneering restructuring and privatization experience of the Chilean ESI, which began in 1980s. Chile has had its share of success and failures over the years and offers an early example of the difficulties of dealing with three distinct segments of the value chain, namely generation, transmission and distribution, along with an independent system operator (ISO). In the Chilean context, power generation was left to competitive investors, transmission was turned into an open–access regime, and distribution was left as a regulated natural monopoly. The system was tested during a severe drought in the late 1990s and subsequent crises, including chronic shortages of natural gas from neighboring Argentina. After 25 years of fine-tuning, the electricity market in Chile is still far from perfect, and because of the diverse and extreme conditions that it has experienced and what has been done, it offers useful insights for policy-makers and regulators around the world. Among the key lessons that can be learned from the Chilean experience are the importance of flexible prices, demand-side management tools, ancillary services, diversified sources of energy, binding international bilateral agreements and the constraints of modeling assumptions. In Chapter 4, David Newbery describes the experience of Britain, another pioneering market from which much has been learned, and continues to be learned. The British market may be characterized as the exemplar of electricity market reform, demonstrating the importance of ownership unbundling and workable competition in generation and supply. The original privatization created a de facto duopoly that supported increasing price–cost margins and induced excessive entry. This problem was addressed by trading horizontal for vertical integration in subsequent mergers. Competition arrived just before the Pool was replaced by the New Electricity Trading Arrangements (NETA). NETA, which intended to address many of the claimed shortcomings of the Pool, however, has had ambiguous market impacts. Increased competition caused prices to fall, inducing generators to withdraw capacity from the market and resulting in concerns about security of supply. Subsequently, price–cost margins increased and capacity was returned to the market. NETA was extended to Scotland in 2005 as the British Electricity Trading and Transmission Arrangements (BETTA) and the entire British transmission system is now under a single system operator. In Chapter 5, Eirik Amundsen, Nils-Henrik von der Fehr and Lars Bergman provide an overview of the evolution and subsequent expansion of the Nordic market, considered among the most successful competitive markets in the world. The Nordic market was put to a severe test during the drought of 2002–2003 where reservoirs in the hydro-dominated system fell to unprecedented low levels. But despite this natural shock, the market held together, without mandatory rationing, blackouts, price manipulation or major financial ruin of any of the players. This, in contrast to other markets – notably California – is an important hallmark of the Nordic market. In Chapter 6, Alan Moran describes how the Australian electricity market drew from the experiences of the UK and has forged a national electricity market (NEM) out of the five interconnected states. The NEM has proven effective in providing adequate incentives to encourage appropriate levels of new investment and the increased competition has driven prices down. Part of the system, which prior to 1994 was exclusively government owned, is now privatized, while some states have maintained government ownership. Despite shortcomings

Why Restructure Electricity Markets?

45

of government ownership, increased efficiency is observable throughout the industry; the available data does however indicate that the private businesses have lower costs and better reliability. Future risks include government intervention, which might discourage optimal investment decisions, may increase costs and could threaten reliability. A major policy question has concerned transmission and how to ensure that it is built according to an incentive structure that is comparable to that facing generation. Some maintain that there is inadequate transmission between major state systems while others argue that there is over-build centered on some state governments’ wishes to export electricity without their generators facing the full costs. Australia’s long, stringy transmission network throws into greater relief issues seen elsewhere namely that new capacity and increased transmission can be substitutes and if one is favored over the other this can lead to poor investment choices and eventually cause supply shortfalls. In Chapter 7, Geoff Bertram traces New Zealand’s failure over two decades of reform to establish a viable industry self-governance framework, and the parallel failure to achieve restraint on monopoly profits by means of light-handed regulation. Starting from a classic monopoly of generation, transmission, distribution and retailing, New Zealand corporatized the ESI, separated lines businesses from generation and retail, removed retail franchises, and broke up the monopoly generator into five companies. These measures were insufficient to achieve competitive outcomes in the absence of hands-on regulation. Generators integrated vertically by takeover of retailers, and the resulting retail oligopoly erected an effective barrier to entry by withholding affiliated generators’ capacity from the very thin market for hedge contacts. Grid pricing and contract provisions foreclosed demand-side innovation and distributed generation. Distribution lines businesses ramped up markups from 30% to 70% without any regulatory restraint and were allowed to revalue their assets to underwrite the new high margins. Faced with failure of the original design, the Government in 2003 established a new industry regulator and invested in a new state-owned thermal generation to plug the country’s yawning gap in reserve generating capacity. In Chapter 8, Gert Brunekreeft and Dierk Bauknecht provide an overview of the German electricity market, still confronting major obstacles. The authors examine the problems and prospects for, and consequences of, new investment in the German market. The potential ramifications of a variety of energy policies are examined in the context of three important energy policy goals: preservation of the environment, provision of supply adequacy and fostering competition. While many countries are concerned about the adequacy of generation capacity, Germany follows a hands-off approach. To promote competition, Germany has now translated the latest European directive into national law, creating a specific regulator for the ESI. Environmental policy has a strong focus on promoting renewables, outside of the electricity market. The new European emissions trading scheme is shown to have a strong influence on future development. In Chapter 9, Reinhard Haas, Jean-Michel Glachant, Nenad Keseric and Yannick Perez cover the evolving situation in the key continental European markets, including Austria, Belgium, Czech Republic, France, Germany, Hungary, Luxemburg, The Netherlands, Poland, Portugal, Slovenia, Slovakia, Spain and Switzerland. Due to its geographic size, multiplicity of national regulatory agencies and trans-boundary transmission issues, attempts by the European Union to create a fully integrated market is progressing slowly. The authors identify a number of necessary conditions, not yet fully achieved, for a wellfunctioning competitive market including separation of grid from generation and supply, sufficient transmission and generation capacity, significant number of generators and a fully liberalized market. Failure to meet these conditions will result in distorted markets with suboptimal results.

46

Electricity Market Reform

In Chapter 10, James Sweeney provides an overview of California’s electricity restructuring in mid-1990s. California’s goal was to deregulate both the wholesale and the retail electricity market, after a transition period during which utilities would be able to recover stranded costs. Three regulatory rules were in place during the transition period: (a) regulators forced investor-owned utilities (IOUs) to divest at least 50% of their generating assets; (b) regulators precluded the utilities from entering forward contracts to acquire electricity; and (c) regulators imposed retail price caps on the utilities. In the midst of the transition period, the Western electricity crisis, driven primarily by a sharp reduction in available hydropower and limitations on natural gas supplies, drove wholesale electricity prices throughout the western USA to triple-digit or quadruple-digit levels. The three regulatory rules together proved disastrous, driving one utility into bankruptcy and another near bankruptcy, and left the state with billions of dollars of losses. Flaws in the California market design encouraged market gaming or exercise of market power by generators, traders, utilities and the state itself. With the crisis over, California has now established resource adequacy rules for utility electricity acquisition, encourages long-term contracts, and has eliminated retail price caps. An ISO-led market redesign project is underway to restructure transmission congestion management. Retail deregulation has been temporarily put on hold. And the future regulatory direction remains uncertain, with some advocates trying to force a return to a vertically integrated monopoly structure and others striving to improve regulatory rules within a partially regulated, partially deregulated system. In Chapter 11, Parviz Adib and Jay Zarnikau describe the much more successful experience of the Texas market, widely acknowledged as the most “robust” in North America to date. The Electric Reliability Council of Texas (ERCOT) has benefited from a confluence of positive factors, including a phased approach where the restructuring of the wholesale market preceded the retail choice, ample generating capacity at the outset of retail competition, and an intra-state market where a single state-level regulatory authority wielded near-exclusive jurisdiction over the implementation of the state’s restructuring plan. A number of challenges, however, remain as Texas seeks to implement more efficient means of managing transmission congestion and ensuring resources adequacy. In Chapter 12, Michael Trebilcock and Roy Hrab provide an overview of the Canadian market, pointing out that, so far, only limited restructuring has occurred. The chapter focuses on restructuring initiatives in Canada’s most populous province, Ontario, including initial setbacks and ramifications of re-regulation, and the lessons learned. The restructuring experience of the province of Alberta is also briefly examined. Both provinces altered their restructuring plans after experiencing unexpected price increases. In particular, the Ontario experience illustrates the importance of political commitment and how restructuring policies can be reversed quickly when a government fears a political backlash. In Chapter 13, Joe Bowring provides an overview of the PJM, considered by many as a role model for an efficient, centrally managed electricity market. The PJM interconnection manages the largest centrally dispatched control area in North America and operates the largest competitive wholesale electricity market in the world. PJM operates a bid based, security constrained, economically dispatched, locationally priced market with open–access transmission and financial transmission rights. PJM has achieved its success to date based on a number of factors including considerable experience as a power pool, well-defined market rules, a workable governance structure and an independent market monitoring function. The PJM model offers useful lessons for other aspiring ISOs and transmission system operators (TSOs). In Chapter 14, Richard O’Neill, Udi Helman, Benjamin Hobbs and Ross Baldick provide a background on important decisions emanating from the Federal Energy Regulatory Commission

Why Restructure Electricity Markets?

47

(FERC) and some of the frustrations the agency has confronted as it attempts to implement its vision of an efficient national market with standardized rules and protocols. While Chapters 10, 11 and 13 examine particular markets with an ISO, this chapter provides a cross-cutting perspective on all the US ISO markets and FERCs broader public policy agenda. The authors provide perspective on regulatory reforms in the USA that began the movement toward organized, competitive markets. These reforms were accompanied by intense debate over the appropriate design of such markets and whether and how to mitigate market power. The chapter examines what motivated these debates and how different ISO regions experimented with alternative approaches. It explains why certain design features eventually were considered best practices and what issues remain for design consideration. It examines what measures of market performance have been implemented and how they should be interpreted. Finally, the chapter provides a summary view of the state of market design and market performance across each US ISO market. In Chapter 15, Taff Tschamler reviews the decidedly mixed experience in the US retail market. Despite a few state-level success stories, a national competitive US retail market has failed to materialize and does not even appear to be in the cards. This chapter provides an overview of the USA experience with competitive retail power market design with focus on default generation service. It sets forth the author’s assessment of effective default service policy, assuming an objective of vigorous competition and customer participation in retail electric markets. It also provides a summary of the various default service approaches or models adopted by states that allow customer choice. This chapter also provides historical data on the level of market activity by jurisdiction. In Chapter 16, Joao Lizardo R. Hermes de Araujo describes rather unique features of Brazil’s large hydro-based power system. A striking feature of market liberalization, which followed the debt crisis of 1980s, was that the divestment process and market reform followed two parallel and nearly independent paths. This, plus the mismanagement of reform and transition – particularly in view of the difficulties found in privatizing large generators – led to the 2001 power crisis. The new arrangement instituted by the Lula administration purports to ensure adequate expansion investment with an expanded role for central planning and coordination including a mechanism for regulated expansion auctions and contracts. There also is a role for contracting for the short and middle terms, which may grow. In principle, the new arrangements could solve the conundrum of thermal power investment in a large hydrodominated system. There are some positive signs that investment in transmission appears to be working. However, two major issues remain to be addressed: streamlining of environmental licensing of hydro plants, and how to ensure an adequate supply of gas to thermal plants and build the gas network in a market-oriented context. In Chapter 17, Isaac Dyner, Santiago Arango and Erik Larsen describe the market reform experience in Argentina and Colombia, not the first in Latin America, but presently considered among the most successful. This verdict, however, might change in the near future. Both are developing nations with a large hydroelectricity generation base with many similarities, yet followed different approaches to market reform more than 10 years ago. The elapsed time provides an opportunity to compare and contrast how well each market has fared and yield insights into the consequences of the initial and subsequent decisions on the market design. Together, the four Latin American countries covered in this volume offer a rich opportunity to examine the effects of different approaches to market liberalization. In Chapter 18, Mika Goto and Masayuki Yajima describe the early experience of introducing limited competition in the Japanese electricity market.

48

Electricity Market Reform

Japan started the market liberalization process in 1995 with the objective of improving the efficiency of vertically integrated monopoly power companies. While some other countries have taken bold – sometimes hasty – steps, Japanese policy-makers are taking a methodical and judicious approach. This, the authors argue, provides an opportunity to incorporate lessons from experiences of other countries, avoiding some of the pitfalls. This chapter reviews the ongoing debates about the vertical separation of utility companies, introduction of retail competition in the residential sector, dilemmas on how to preserve the country’s strategically important nuclear industry and insure adequate investment in infrastructure.

References Averch, H. and Johnson, L.L. (1962). Behavior of the firm under regulatory constraint. American Economic Review, 52, 1053–1069. Jamasb, T. and Pollitt, M. (2005). Electricity market reform in the European Union: review of progress towards liberalization and integration. The Energy Journal Special Issue on European Electricity Liberalization, 11–41. Kollmann, A. et al. (2005). Netztafiein Österreich: Bestandsaufnahme und Internationaler Vergleich, Linz. Littlechild, S. (2005). Beyond Regulation, IEA/LBS Beesley Lectures on Regulation Series XV. 4 October 2005, forthcoming in volume edited by C. Robinson, published by Edward Elgar. Michaels, R. (2005). Rethinking vertical integration in electricity. Department of Economics, California State Fullerton, Fullerton, CA, May. Newbery, D. (1999). Privatization, Restructuring, and Regulation of Network Utilities. MIT Press, Cambridge, MA. p. 148. Pollitt, M.G. (1997). The restructuring and privatization of the electricity supply industry in Northern Ireland – Will it be worth it? Mimeo, Cambridge University, February.

Chapter 2 Sector-Specific Market Power Regulation versus General Competition Law: Criteria for Judging Competitive versus Regulated Markets GÜNTER KNIEPS Institute for Transport Networks and Regional Policy, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany

2.1. Introduction Since the comprehensive abolishment of legal barriers to entry into (almost) all network sectors, network economics has experienced a paradigm shift. Before the opening of the markets the controversial question was if and to what degree competition in networks could function at all; meanwhile, the central controversy of network economics has shifted to the distribution of tasks between sector-specific regulation and general competition law. The application of sector-specific regulatory intervention constitutes a massive intervention in the market process which requires a particularly well-founded justification. That the general competition law should be applied to the opened network sectors, too, is beyond dispute. Sector-specific regulatory interventions with competition policy objectives, on the other hand, are only justified if there is network-specific market power.1 Insofar as vague legal terms originating from general competition law – such as, for instance, market dominance – are being used to determine the need for sector-specific intervention they have to be corroborated by a localization of market power that is substantiated by network economics. Otherwise it is to be expected that market power is simply postulated, but not actually localized. A suitable economic reference model for establishing the regulatory activity necessary for disciplining market power in network sectors must be able to take into account the essential characteristics of networks without automatically equating them with market power. In the debate on the potentials of competition in opened network sectors the theory of contestable markets, developed by American economists in the 1970s (e.g. Baumol et al., 1982), 1

Technical regulatory functions (network security, allocation of frequencies, number administration, etc.) and the pursuit of universal service objectives by means of entry-compatible instruments (e.g. universal service fund) also constitute long-term sector-specific regulatory tasks, but are not examined in detail here.

49

50

Electricity Market Reform

is of central importance. This theory formulates the conditions that have to apply for potential competition to function as perfect substitute for active competition – under the assumption that there is a natural monopoly situation where one supplier can serve the market at lower cost than several suppliers. Critics of the theory of contestable markets refer in particular to the non-robustness of this theory (cf. e.g. Schwartz and Reynolds, 1983; Schwartz, 1986). Even the smallest alterations in the assumptions of the theory of contestable markets would destroy the disciplinary effects of potential competition and thus, it is argued, this concept is largely irrelevant for regulatory policy. As a consequence of this non-robustness of the theory of contestable markets the necessity of global sector-specific regulation may be postulated by the proponents of regulation. In the following it is shown that while the theory of contestable markets formulates sufficient conditions for the absence of network-specific market power, it does not provide a comprehensive definition of the preconditions for the existence of network-specific market power and the resultant need for sector-specific regulation. Whereas the theory of contestable markets examines only the role of potential competition with identical cost functions for both active providers and potential competitors (cf. Panzar and Willig, 1977; Baumol, 1982), effective network competition is not limited to potential competition. Newcomers can successfully distinguish themselves from the incumbent and find their own niche in the market, in particular by means of technological and product differentiation. An essential contribution of the theory of contestable markets to the study of potential competition consists in basing its analysis on Stigler’s concept of a barrier to entry (Stigler, 1968). For Stigler, economies of scale do not constitute barriers to entry, as they do not cause long-term asymmetries between incumbent and potential competitors. At this point, the theory of monopolistic bottlenecks, which is centrally concerned with localizing stable network-specific market power, becomes relevant. Based on this theory, a disaggregated regulatory policy can enable effective competition on complementary network parts by means of suitable access and price regulation. In this context, the essential facilities doctrine, which originates from US antitrust law, is taken up and generalized as a rule for the class of monopolistic bottlenecks. A monopolistic bottleneck in need of regulation only exists when both potential and active competition are lacking. In addition to potential competition, it is therefore also necessary to take into account the various potentials of active competition by means of product differentiation, price differentiation, technological differentiation, etc. As the theory of contestable markets disregards active competition, it is incapable of providing a comprehensive description of the potentials of competition after network opening. The theory of monopolistic bottlenecks was not specifically developed for a particular network sector, but is rather an economically sound instrument for localizing and disciplining remaining network-specific market power in all network sectors (e.g. railways, air traffic, telecommunications, etc.). It is shown that network-specific market power can be localized in most network industries, although strongly varying in extent among different network industries. In particular, monopolistic bottlenecks are relevant in electricity grids (transmission and distribution). Gas distribution pipelines are monopolistic bottlenecks too, although active network competition in supra-regional transmission pipelines may occur. Insofar as there are monopolistic bottleneck areas in network sectors, they require sectorspecific regulation in order to discipline the remaining market power. In particular, symmetric access to monopolistic bottleneck areas for all active and potential suppliers of network services must be guaranteed, so that competition can become fully effective on all complementary markets. It is shown that the application of regulatory instruments should be limited to the monopolistic bottleneck basis. Price-cap regulation should be applied, leaving the design of flexible pricing structures for network access to individual firms. Although

Market Power Regulation versus General Competition Law

51

perfect regulatory instruments do not exist, neither end-to-end regulation, nor access holiday regulation can be recommended. This chapter is organized as follows: Section 2.2 is devoted to an analytical foundation of the concept of network-specific market power. After an outline of the theory of contestable markets there follows a short description of the theory of monopolistic bottlenecks and its relation to the essential facilities concept. Section 2.3 contrasts the theory of imperfectly contestable markets with the theory of monopolistic bottlenecks. It is shown that the controversy about the relevance of the theory of contestable markets, in particular the criticism of non-robustness, does not constitute a fundamental argument against the importance of localizing network-specific market power. On the contrary, it is shown that monopolistic bottlenecks are robust structural characteristics that exist independently of alternative behavior assumptions. In Section 2.4 a short survey of the localization of monopolistic bottlenecks in different network industries is provided, leaving a more explicit analysis of localizing monopolistic bottlenecks within the energy sector to Section 2.5. In Section 2.6 the concept of a disaggregated regulation of monopolistic bottlenecks is provided, avoiding end-to-end regulation as well as access holidays.

2.2. The Search for Sector-Specific Market Power 2.2.1. The theory of contestable markets The concept of contestable markets takes its starting point from the question of the disciplinary effects of potential competition in natural monopoly areas.2 The threat of potential market entry was recognized as early as 1859 when Chadwick pointed out the difference between “competition for the field” and “competition within the field of service”. In 1968 Demsetz proved that the existence of a natural monopoly does not per se necessitate regulation. The almost axiomatic connection between the production structure of a natural monopoly and its need for regulation with respect to market entry, market exit, and pricing in a world without uncertainty does therefore not stand up to economic analysis. Auctioning off the right to supply a market that is a natural monopoly can, if necessary, replace competition in the market. According to Demsetz (1968, p. 58) the following two conditions are crucial for the functioning of such an auctioning process: competition on the input markets (many potential bidders), and no prior arrangements between competing bidders. In order to define the effects of the threat of potential competition more precisely, in the second half of the 1970s the theory of contestable markets was developed (Baumol, 1977, 1982; Panzar and Willig, 1977; Baumol et al., 1982). A market is termed contestable if entry is completely free and exit is completely free of charge (cf. Baumol, 1982, p. 3). The term “free entry” is here characterized by the absence of barriers to entry as defined by Stigler (cf. Bailey and Panzar, 1981, p. 128). However, this does not mean that market entry is free of charge or easy, but rather that new entrants have no cost disadvantages compared to the active supplier. As long as the inputs are available to both active and potential market participants under identical conditions, according to Stigler they do not produce barriers to entry.3 Thus economies 2

A natural monopoly exists if the cost function in the relevant area of demand is subadditive (cf. Baumol, 1977, p. 810). When examining the cost side of networks, economies of scale and scope in providing services, are of particular importance, which can result in a single supplier being able to serve the market at a lower cost than several suppliers. 3 Stigler defines a barrier to entry as “… a cost of producing (at some or every rate of output) which must be borne by a firm which seeks to enter an industry but is not borne by firms already in the industry” (Stigler, 1968, p. 67).

52

Electricity Market Reform

of scale do not cause a barrier to entry, as long as the entrant has access to the same cost function. Stigler’s concept further implies that traditional competition parameters such as product differentiation and the resultant creation of reputation and goodwill or the requisite capital do not constitute barriers to entry either, as they, too, affect active and potential competitors alike. In other words, in these contexts the cost function is only dependent on those factors to which all (potential and active) market players have symmetrical access.4 The theory of contestable markets proposes a model framework where, in the case of natural monopolies, potential competition is the perfect substitute for absent actual competition provided by active market players. The following assumptions are being made: ●

●

●

Free market entry: There is a large number of potential competitors with unlimited access to the most cost-efficient technology of a natural monopoly. Absence of irreversible costs:5 The investment necessary for market entry can be fully recovered in case of market exit. Exit is possible without incurring costs or loss of time. Bertrand–Nash behavior: The potential competitors calculate their chances on the market by taking the incumbent’s current prices as given and undercutting them. Perfect information on the part of all players is assumed, that is, there are no searching costs, so that even small changes in prices result in an immediate switch of the total demand to the cheapest supplier. Customers do not feel bound to a specific firm; they can switch with little effort when a new supplier appears.

The essential characteristic of a contestable market is its susceptibility to hit-and-run entry. Even a small, temporary profit on the part of the active supplier creates incentives for the entry of a potential competitor. Prerequisite for this, however, is a sufficient flexibility in prices, so that a potential entrant can undercut the active supplier (cf. Bailey, 1981, p. 178). The question may arise if and to what extent the basic assumptions of the theory of contestable markets are relevant in the real world at all. It is indeed an essential characteristic of the functions of competition on the open markets for network services that business strategies such as product differentiation, price differentiation, creation of goodwill, creation of an efficient distribution network, etc. are also used strategically. In addition, information problems (search costs, asymmetric information, etc.) can be relevant.6 It is true that the 4

The barriers to entry that traditional industrial economics – based on Bain (1956) – distinguishes (economies of scale, product differentiation, high requirements of capital), on the other hand, do not allow a reliable proof of stable market power (cf. e.g. Schmalensee, 1989). For instance, von Weizsäcker (1980a, b) shows that reputation and goodwill constitute efficient mechanisms for reducing uncertainty, and thus may lead to an increase in social welfare. 5 For the incumbents, irreversible costs no longer affect decision-making. Potential entrants on the other hand have to decide whether or not to incur these irreversible costs in the market they wish to enter. The incumbents, therefore, have lower decision-relevant costs than the potential entrants. Irreversible costs in combination with a natural monopoly constitute a credible threat that may discourage a second network operator from entering the market. Although even the irreversible costs have to achieve riskequivalent rates of return, they would be irrevocably lost after market entry. Therefore, the threat that the incumbent could temporarily reduce its prices to the variable cost level is indeed credible. 6 The Bertrand–Nash assumption of the theory of contestable markets does not aim to deny the more or less severe information problems of real markets. No sector-specific stable market power can be derived from information problems alone, because markets are resourceful where the (endogenous) development of institutions for overcoming information problems is concerned, for instance by the creation of goodwill. On the other hand, in natural monopolies with irreversible costs stable market power exists, even if all players are perfectly informed.

Market Power Regulation versus General Competition Law

53

simplified models employed by the theory of contestable markets do not provide a comprehensive description of the manifold potentials of competition. However, this fact should not lead to the opposite conclusion that competition will therefore not work on principle. Neither does this mean that the application of general competition law on these markets should be rejected. However, as on all other competitive markets, the burden of proof whether market power exists and whether it is being abused rests with the competition authorities. In contrast to a general sector-specific regulation, such interventions into the competitive process should only be made case-by-case and ex post.7

2.2.2. The theory of monopolistic bottlenecks 2.2.2.1. Localizing network-specific market power The theory of monopolistic bottlenecks is based on a strict application of Stigler’s concept of barriers to entry in order to identify network-specific market power. It constitutes a central component of the disaggregated regulatory approach (cf. Knieps, 1997a, pp. 327ff.; 1997b, pp. 362ff.): the localization of network-specific market power in order to determine the minimal regulatory basis. Its objective is to derive, based on the principles of network economics, a regulatory basis consistent for all network sectors which justifies sector-specific regulatory inventions, irrespective of historical or institutional coincidence. All other network areas are subject to general competition law. The special focus of regulatory activity should be on the design of a symmetrical regulation of the access to monopolistic bottlenecks, combined with a regulation of access charges. The issue of the monopolistic bottlenecks, especially the concomitant problem of network access, is frequently discussed in network economics (Baumol and Willig, 1999, p. 44; Laffont and Tirole, 2000, p. 98; Kuhlmann and Vogelsang, 2005, p. 34). The conditions for a monopolistic bottleneck facility are fulfilled: 1. If the facility is necessary for reaching consumers, that is, if no second or third such facility exists, that is, if there is no active substitute available. This is the case if there is, due to economies of scale and economies of scope, a natural monopoly situation, so that one supplier can provide this facility at a lesser cost than several suppliers. 2. If at the same time the facility cannot be duplicated in an economically feasible way, that is, if no potential substitute is available. This is the case if the costs of the facility are irreversible. The entire value chain has to be examined in a disaggregated manner that is it has to be differentiated into those network areas that do have bottleneck characteristics and those that do not. Non-bottleneck areas are characterized by effective competition. The latter is by no means confined to potential competition. Both active and potential competition with and without technological differentiation as well as product differentiation and innovation (of both products and processes) constitute potential parameters of effective competition. Service networks due to the absence of irreversible costs unquestionably have non-bottleneck character; they may or may not possess the characteristics of a natural monopoly. When establishing 7

Competition authorities have to weigh up two possible sources of error. A false positive error occurs, when the competition authority intervenes in the competitive process, even though competition is functioning and there is no need for any active competition policy. A false negative error occurs when the competition authority fails to act, even though competition policy measures are indeed called for. Within the context of analyzing predatory pricing this differentiation has already been pointed out by Joskow and Kleworick (1979).

54

Electricity Market Reform

proof that a facility is a monopolistic bottleneck, it is crucial to concentrate exclusively on those network areas where there is a lack of active as well as potential competition and, consequently, no economically feasible alternative network access on the downstream markets. For instance, if a service network provider can choose between alternative network infrastructure providers, there is no monopolistic bottleneck, even if the infrastructures in question are not identical but, in accordance with the theory of monopolistic competition, characterized by product/technology differentiation.8 In network areas without irreversible costs no stable market power can be localized. Nevertheless fixed costs and economies of scale play an important role in these areas. Weitzman (1983) argues that the natural monopoly problem can only occur when irreversible costs are involved. There are certainly cases conceivable where market entry and exit at no charge is linked to small or non-existing economies of scale; in network sectors, however, economies of scale are usually of crucial importance, even in the absence of irreversible cost.9 Network areas where economies of scale are relevant but sunk costs do not occur are areas where competition is basically effective. The central concern of the theory of monopolistic bottlenecks is the differentiation, in the context of the disaggregated regulatory approach, between those network areas where competition is effective and those in need of being regulated, which are characterized as monopolistic bottlenecks. 2.2.2.2. Monopolistic bottlenecks and the concept of the essential facility When applying competition rules in order to discipline network-specific market power, the concept of the essential facility is of crucial importance. A facility or infrastructure is termed essential if it simultaneously: ● ● ●

is indispensable for reaching consumers and/or for enabling competitors to do business; is not otherwise available on the market; objectively cannot be duplicated by reasonable economic means.

This concept suggests the connection to the essential facilities doctrine, derived from US antitrust law, which is meanwhile being increasingly applied in European competition law also (cf. e.g. Lipsky and Sidak, 1999). The doctrine states that a facility is only to be regarded as essential if the following conditions are fulfilled: entry to the complementary market is not effectively possible without access to this facility; it is not possible for a supplier on a complementary market to duplicate this facility at a reasonable expense,10 and there are also no substitutes (Areeda and Hovenkamp, 1988).11 In the context of the disaggregated regulatory approach the essential facilities doctrine is no longer applied case by case – as is common in US antitrust law – but to an entire class of cases, namely, monopolistic bottleneck facilities. The design of non-discriminatory conditions of 8 However, the absence of stable network-specific market power when there are two or more infrastructure suppliers present does not mean that practical competition policy (e.g. merger control) is redundant. 9 For this cf. also Baumol (1996), pp. 57f. 10 Thus it is not feasible to offer, for instance, a ferry service without access to ports. 11 Occasionally an additional criterion for applying the essential facilities doctrine is formulated, namely, that the use of the facility is essential for competition on the complementary market, because it reduces prices or increases supply on this market. This criterion, however, merely describes the effects of access.

Market Power Regulation versus General Competition Law

55

access to essential facilities must be specified in the context of the disaggregated regulatory approach. It is important in this context to view the application of the essential facilities doctrine in a dynamic context. Therefore, an objective for the formulation of access conditions must be to not obstruct infrastructure competition, but rather create incentives for both research and development activities, and innovation and investment on the facility level. 2.2.3. The theory of contestable networks as precursor for the theory of monopolistic bottlenecks The theory of contestable markets was developed at a time when even in the USA the process of opening the network sectors for competition was still in its early stages.12 The question that concerned economic policy was whether competition can function if economies of scale are involved (e.g. Willig, 1980): The theory of contestable markets was focussed on those markets that, due to significant economies of scale, promised, not so much active competition, but rather effective potential competition: for example, air transport or bus transport.13 The competition policy point of reference was the concept of the “invisible hand” of perfect competition, which is formulated in an extreme version by the general equilibrium theory. As the general equilibrium theory disregards the existence of economies of scale (e.g. Debreu, 1959), the theory of contestable markets was to permit them (and the concomitant non-convexities of production technology) in its model analysis, as they are of central importance for network sectors. Even though the theory of contestable markets is conceived as a partial analysis, all the other assumptions of the general equilibrium theory – complete information (no searching costs, no asymmetric information), no externalities, price as a perfect strategy parameter, no product or price differentiation, no dynamics – were transferred to the modeled world. Whereas market entry is not explicitly dealt with in the general equilibrium theory (all firms are seen as “atoms”, their active number determined endogenously in equilibrium), for the theory of contestable markets the role of potential competition is the focal point. The theory of contestable markets states that if there is a network structure with its concomitant economies of scale, competition in the form of potential market entry may quite possibly be effective. It is precisely by formulating the maximum role of potential competition for a scenario with a natural monopoly in combination with an absence of active competition that the theory of contestable markets laid an important cornerstone for the economically well-founded localization of network-specific market power after comprehensive market opening. The theory of monopolistic bottlenecks and the theory of contestable markets have their common origin in Stigler’s concept of a barrier to entry. The focus is therefore on the longrun cost asymmetries between active supplier and potential entrant. Although industrial economy’s quest for the “right” concept of a barrier to entry is still ongoing (cf. Carlton, 2004; McAfee et al., 2004; Schmalensee, 2004), Stigler’s concept has proven to be undoubtedly suitable for localizing stable network-specific market power. Neither adjustment costs nor the evolutionary dimension of market processes detract from the pivotal role of Stigler’s concept of a barrier to entry for localizing network-specific market power. The theory of contestable markets was rightly criticized for, due to its static character, disregarding essential functions 12

Thus interstate aviation was opened for competition in 1978, and interstate telecommunications – voice telephony in 1982, while intrastate network sectors remained almost completely closed. 13 In the USA, market opening for railway transport was not even considered for a long time; for the opening of this sector Europe was the pioneer.

56

Electricity Market Reform

of competition (cf. Carlton, 2004, pp. 468ff.); as regards competitive network sectors, however, the only conclusion to be drawn from this is that the manifold types of active competition are also important. However, it is by no means possible to derive the non-existence of stable network-specific market power by referring to network dynamics. Therefore, a periodic review of the phasing out potentials of monopolistic bottleneck regulation due to newly developed network alternatives would seem to be necessary (cf. Knieps, 1997a). 2.3. Imperfectly Contestable Markets versus Monopolistic Bottlenecks 2.3.1. The focus of the empirical studies on the theory of contestable markets The empirical studies on the role of potential competition examined the performance effects (e.g. price drops) from an end-to-end perspective, that is, on the end consumer markets.14 The effects of potential competition and price and product differentiation (e.g. hub formation, frequent flyer programs) were included, as well as the role of airports with scarce starting and landing capacities. Although the special importance of access to airports was particularly stressed (e.g. Bailey and Panzar, 1981, pp. 132f.), network access problems were analyzed on a level with other market imperfections.15 Two contrary estimates for the changes in performance to be expected on contestable markets can be distinguished. According to Bailey and Baumol (1984, pp. 130ff.), following market opening, the importance of nonoptimal network structures and of cost structures diverging strongly between businesses should decrease and contestability should improve.16 According to Bailey and Williams (1988, p. 191) market opening leads to a considerable increase in competitive strategies, via the price-cost range of the industries. According to Graham et al. (1983, p. 137) “schedule and service reliability” causes asymmetry in favor of the incumbent. This is designated as sunk costs, even though it does not constitute a barrier to entry as defined by Stigler. A disaggregated regulatory approach differentiating systematically between network areas with bottleneck character and those without bottleneck character, and thus dealing with the design of non-discriminatory access to airport capacities separately was not developed at that time. Consequently, there was no analysis of the question to what extent adequate network access conditions create the basis for active and potential competition on service markets in the first place. Thus no sufficient differentiation was possible between, on the one hand, the effects of market imperfections (e.g. information problems) – disregarded by contestability theory, because they do not only occur in network sectors but also on other markets – and, on the other hand, network-specific structural characteristics that need to be regulated, because they systematically impede network access and constitute network-specific market power. 2.3.2. One-sided focus on the role of potential competition At the center of the empirical studies on the theory of contestable markets was the question whether market opening did indeed show the effectiveness of potential competition. Those markets on which performance was influenced not only by potential suppliers but also by 14

Among the large body of literature on this issue (e.g. Bailey and Panzar, 1981; Caves et al., 1984; Morrison and Winston, 1987; Bailey and Williams, 1988). 15 Measuring the effects of market opening empirically is inherently very difficult (cf. Bailey and Panzar, 1981, p. 144). Thus questions arose such as: Is the industry not in balance? What is the role played by fluctuating fuel prices or the cost of intermodal alternatives; or, more generally, how would the market have developed if it had not been opened? 16 Cf. for this also Caves et al. (1984, p. 484).

Market Power Regulation versus General Competition Law

57

the number of active suppliers were termed “imperfectly contestable” – as opposed to “perfectly contestable markets”, on which even a single potential entrant can fulfil a comprehensive disciplinary function (Morrison and Winston, 1987, pp. 58f.). But the disciplinary role of potential competition is also derived within the concept of imperfectly contestable markets (cf. e.g. Morrison and Winston, 1987, p. 61). Deviations from perfect contestability are attributed to, among other things, the importance of switching costs. The theory of contestable markets, on the other hand, is based on Stigler’s concept of barriers to entry and thus on long-run cost asymmetries (cf. Bailey and Panzar, 1981, p. 128) and abstracts from these types of market imperfections. It is conspicuous that in those network sectors where network infrastructures did not cause access problems (road haulage, bus transport, international shipping line services) the contestability hypothesis was hardly controversial (cf. Baumol and Willig, 1986, pp. 24ff.; Davies, 1986). Even though aviation had at first been considered the prime example of a contestable market, the criticism of the theory of contestable markets focussed increasingly on this sector. Even its supporters relativized the hypothesis of perfect contestability in the direction of imperfectly contestable markets.17 In their study on the American air transport market Bailey and Panzar (1981, p. 145) differentiate strictly between the roles of active and potential competition. The authors show that active competition in the air transport sector between long distance and regional providers can discipline the market for regional transport, even if potential competition on the regional level is absent. The implicit conclusion, however, is that on opened markets not only potential but also active competition can fulfil a disciplinary function – in conjunction with a concomitant differentiation of the cost function and product differentiation (small versus large aeroplanes, different times of departure, etc.). The full effect of a comprehensive market opening cannot be registered in its entirety by the theory of contestable markets. The role of access to infrastructures is also shown to be a source for the lack of perfect contestability (Morrison and Winston, 1987, pp. 61f.). Bailey and Panzar (1981, pp. 132f.) also emphasize the relevance of sunk costs for airports. But there is no strict differentiation being made between markets with stable market power (monopolistic bottlenecks) that are in need of regulation and markets with effective competition where sector-specific regulation is redundant. 2.3.3. Imperfect contestability versus monopolistic bottlenecks The studies on imperfect contestability transcend the abstraction level of the theory of contestable markets. Thus searching costs and information problems are taken into account, which may occur on all markets. The role of cost heterogeneity and product differentiation and the concomitant active competition are regarded as an indication for the absence of perfect contestability, although it cannot be concluded from this that there is no effective competition. It has been argued that, in spite of the problems of asymmetric access to infrastructures and the resultant market imperfections, competition functions in the sense of imperfect contestability (Bailey and Panzar, 1981, p. 134; Morrison and Winston, 1987, pp. 55f.). But this view does not obliterate the necessity to determine the remaining need for regulation in opened network sectors and to discipline it by means of suitable regulatory instruments. As has already been shown in Section 2.2.2, the existence of monopolistic bottlenecks and the localization of network-specific market power needing to be regulated presupposes the 17

It should also be noted that the contestability hypothesis was not tested for the American railway sector. AMTRAK still holds the monopoly for railway passenger transport services.

58

Electricity Market Reform

lack of both active and potential competition. If, following market opening, both active and potential competition emerge, the objective of creating effective competition is fulfilled. The differentiation between, on the one hand, markets on which only potential competition can be observed and, on the other hand, markets where there is both potential and active competition is irrelevant for the localization of monopolistic bottlenecks. 2.3.4. On the independence of network-specific market power from alternative behavioral assumptions The theory of contestable markets is a static theory (cf. Baumol et al., 1983, p. 495). It describes, by means of the Bertrand–Nash behavioral assumption, the maximum efficiency of potential competition in natural monopoly areas without irreversible costs. Here even minor price reduction potentials lead directly to market entry, so that residual demand is perfectly elastic (cf. Knieps and Vogelsang, 1982, p. 236). It has to be kept in mind that what is concerned here is a static equilibrium concept. Behavioral assumptions about dynamic qualities are irrelevant for determining the Bertrand–Nash equilibrium. In this sense, the non-robustness criticism by Schwartz and Reynolds (1983, p. 488) is misleading, because it is based on assumptions about dynamic qualities of the relation between entry lag and price adjustment lag,18 which are unnecessary for the concept of perfectly contestable markets and do not constitute a logical consequence of the absence of irreversible costs either (cf. Baumol et al., 1983, p. 496). Beside the Bertrand–Nash behavior assumption, however, other behavioral assumptions (e.g. the Cournot–Nash or von Stackelberg behavioral assumptions) are conceivable which could impede the disciplinary effects of potential competition considerably, even if there are no irreversible costs involved (e.g. Knieps and Vogelsang, 1982, pp. 236ff.). Basically, what is concerned here are quantity pre-commitments, whereas the theory of contestable markets emphasizes the formulation of pricing behaviors without quantity pre-commitments (Baumol and Willig, 1986, p. 14). Although such quantity pre-commitments cannot be completely excluded in markets without irreversible costs – just like in all other markets – no networkspecific stable market power that would justify sector-specific regulation can be derived from this fact. In the absence of irreversible costs, network-specific market power that would be robust under alternative behavioral assumptions cannot be proven, even in the case of a natural monopoly (cf. Knieps and Vogelsang, 1982). Market power on the basis of the Cournot–Nash behavior assumption becomes immediately unstable under the Bertrand– Nash behavior assumption. Interventions by competition authorities would thus have to be based on behavior hypotheses that are difficult to verify empirically. Even according to the Bertrand–Nash behavioral assumption, with the demand side perfectly willing to switch and perfectly informed, its exploitation by the supplier of a monopolistic bottleneck facility cannot be prevented. This is all the more true if the customers of a monopolistic bottleneck facility have tied themselves to a specific infrastructure provider via pre-commitments (e.g. in the context of the Cournot–Nash behavioral assumption), even if market entry cannot be excluded for all scenarios (cf. Knieps and Vogelsang, 1982, pp. 239f.). As a consequence, in the absence of irreversible costs network-specific market power does not occur even if the behavioral assumptions are altered. The reverse is also true, that is, the 18

While the entry lag mentioned above refers to the time period between the entry of a newcomer and the time when the entrant is able to sell his product, the price adjustment lag indicates the period between market entry and the incumbent’s price reaction.

Market Power Regulation versus General Competition Law

59

structural characteristics of a natural monopoly in combination with irreversible costs create market power, which remains stable even under alternative behavioral assumptions. 2.3.5. The “␧ sunk cost” argument versus the localization of network-specific market power The effectiveness of potential competition – and thus the central statement of the theory of contestable markets – was challenged with the argument of non-robustness, that is, that even the existence of minimal irreversible costs could lead to a non-contestable monopoly scenario (“␧ sunk cost” argument) (cf. Schwartz and Reynolds, 1983; Schwartz, 1986, pp. 41ff.). Baumol et al. (1983) do not completely exclude this possibility, but on their part point out conditions under which, with very low irreversible costs, the performance of the markets approximates that of a perfectly contestable market. The most elementary formulation of this argument of the non-robustness of the theory of contestable markets in case of minimal irreversible costs can be found in Stiglitz (1987, p. 889). Stiglitz assumes that minimal irreversible costs can occur even on service markets and with almost all investments; he is, however, using a broader interpretation of the term irreversible costs than Stigler in his concept of barriers to entry.19 The non-robustness argument that even minimal irreversible costs result in market power is here presented in the context of a subgame perfect equilibrium. The crucial precondition for this is the assumption that all potential entrants are identical to the incumbent, so that it is only the time asymmetry of incurring the minimal irreversible costs that constitutes the potential for a plausible threat on the part of the incumbent. In the framework of this model, newcomers have no possibility to distinguish themselves from the incumbent by means of technology, product or price differentiation. It is immediately evident that the manifold forms of active and potential competition on service markets cannot be comprehensively dealt with by this approach. As regards the problem of localizing network-specific market power, however, the controversy about the role of minimal irreversible costs is misleading. Active network competition typically goes together with product and technology differentiation, so that the existence of minimal irreversible costs does not lead to a plausible threat regarding market entry. Instead, it is the presence or absence of a monopolistic bottleneck that constitutes the crucial factor in deciding whether network-specific market power exists. Only a natural monopoly in combination with irreversible costs and the resultant lack of active and potential competition lead to a scenario in which the non-discriminatory access of all players on downstream markets must be guaranteed by a suitable form of sector-specific regulation.

2.4. Localizing Monopolistic Bottlenecks in Different Network Sectors The theory of monopolistic bottlenecks was not specifically developed for a particular network sector, but is rather an economically sound instrument for localizing and disciplining remaining network-specific market power in all network sectors (e.g. railways, air traffic, telecommunications, etc.). The character and extent of monopolistic bottleneck areas vary considerably between the individual network sectors. Individual proof as to which network area does indeed meet the criteria for a monopolistic bottleneck is necessary, and it is also important to avoid the danger of an erroneous identification of a monopolistic bottleneck. 19 “

While much of investment is not sunk, however, there is a sunk cost element in almost all investments. An airline must advertise to obtain customers; it must solve complicated routing problems” (Stiglitz, 1987, p. 889).

60

Electricity Market Reform Table 2.1. Airports as monopolistic bottleneck facilities.

Air traffic Air traffic control Airports

Natural monopoly

Irreversible costs

X/— X X

— — X

Table 2.2. Railway infrastructure as a monopolistic bottleneck facility.

Railway traffic Railway traffic control Railway infrastructure

Natural monopoly

Irreversible costs

X/— X X

— — X

Table 2.3. Local telecommunications networks as monopolistic bottleneck facilities.

Terminal equipment Telecommunications services (including voice telephone services) Satellite/mobile networks Long-distance cable-based networks Local cable-based networks

Natural monopoly

Irreversible costs

— —

— —

X — X/—

— X X

If, for instance, due to technological progress, the conditions for a monopolistic bottleneck no longer exist, the corresponding sector-specific regulation must also stop (cf. Knieps, 1997a). In an environment of competing network infrastructure providers and a variety of networks and technologies this means that existing regulation should be reduced rather than extended. The combination of natural monopoly and irreversible costs can occur in different network sectors: For example, airport infrastructures – in contrast to aeroplanes – are associated with irreversible costs. Once made, investments in terminals and runways cannot be transferred to another location, the way an aeroplane can. Thus to the extent airports are natural monopolies, they constitute monopolistic bottlenecks. Railway infrastructure, unlike rail transport services and railway traffic control, represents a bottleneck facility, because the track operator holds a natural monopoly and the building of rail tracks involves irreversible costs.20 Air traffic and railway traffic may or may not possess the characteristics of a natural monopoly (depending on the relevant market). In the telecommunications sector, bottleneck facilities can (if at all) only be found in the local loop, while in long-distance networks there is both active and potential competition (Knieps, 2004). Although legal entry barriers do still exist in some parts of letter mail conveyance, postal services as a whole do not constitute a monopolistic bottleneck. The different components of postal services may or may not possess the characteristics of natural monopolies (Knieps, 2002a). Tables 2.1–2.4 illustrate the application of the theory of monopolistic bottlenecks to the network sectors mentioned above.21 20

For a more explicite analysis of the role of monopolistic bottlenecks in the area of transport infrastructures (see Knieps, 2006). 21 For a more detailed discussion, cf. Knieps (1997a); Knieps and Brunekreeft (eds.) (2003).

Market Power Regulation versus General Competition Law

61

Table 2.4. No monopolistic bottlenecks in any component of letter mail conveyance. Components

Natural monopoly

Irreversible costs

Collecting Sorting of outgoing mail Transport Sorting of incoming mail Delivery

X/— X/— — X/— X

— — — — —

2.5. Localizing Monopolistic Bottlenecks Within the Energy Sector 2.5.1. The different value chains in the electricity and gas sectors Within the electricity sector four vertically successive stages can be distinguished: 1. 2. 3. 4.

power generation, transmission grid (high voltage), distribution grid (low voltage), retail.

Within the gas sector the following stages can be distinguished: 1. 2. 3. 4.

extraction (production), supra-regional transmission pipelines, regional and local distribution, retail.

It is important to differentiate between those parts of the value chain, which have the characteristics of monopolistic bottlenecks and the other, competitive parts. The monopolistic bottleneck parts are characterized by stable network-specific market power and should therefore be regulated ex ante. Although the competitive parts are typically not characterized as atomistic markets, the ex post application of competition law is sufficient. 2.5.2. Competitive generation/production and retail It is obvious that electricity generation, gas production as well as the retail level do not possess the characteristics of monopolistic bottlenecks. Several active suppliers can be identified, so that the characteristics of a natural monopoly are not fulfilled. Both traditional industrial economics22 and modern competition theory based on methods of game theory deal quite extensively with problems of market power. However, in many game theory model approaches the market power to be localized is already assumed to exist, and it is on the basis of this premise that the numerous competition-distorting effects of such (assumed) market power are analyzed.23 It can be shown that stable market power in markets with several suppliers cannot be unequivocally proven to exist either by means of empirical/econometric approaches, or by means of game theory approaches, and that, consequently, it is not possible to derive an economically sound basis for ex ante regulatory intervention in oligopoly markets (see appendix). 22 23

For an overview, see Schmalensee (1989), Chapter 16. A good overview of this model world is given in Tirole (1989).

62

Electricity Market Reform

2.5.3. Monopolistic bottlenecks in the transmission and distribution networks 2.5.3.1. The specific characteristics of natural gas transmission and electricity transmission compared For a differentiated analysis of competition potentials to be possible, the specific features of the commodity and the related particularities of transmission have to be considered. As part of such an analysis, it is also very helpful to examine the fundamental differences between natural gas and electricity. Natural gas is a primary energy available as a natural resource. This means that it can only be produced where the gas reservoirs are. Most of these reservoirs are situated abroad. The gas therefore has to be transported over long distances to the centers of consumption. For this reason there is extensive cross-border gas transportation from remote sources. Moreover, there are different gas qualities, which is why customers cannot use just any gas. Natural gas transportation occurs on specific routes between entry and exit points, with the gas in the pipeline flowing predominantly in one direction, even in the long term. An increase in a pipeline’s transportation volume entails an increased pressure drop. This pressure drop describes the difference between the gas pressure at the entry point and the gas pressure at the exit point, or between two precisely defined points of a pipeline (e.g. compressor stations24). Another characteristic feature of gas is the fact that it can be stored. International gas transportation not only relies on pipelines but can also be effected through liquefied natural gas (LNG) tankers. This involves cooling the gas down to its liquid state for transportation, which is then followed by regasification at the point of destination. As a secondary energy, electricity can, in principle, be produced anywhere (subject to regulations and approvals). It is predominantly generated in the country in which it is consumed. Both the location of the power stations and the design of the transmission network are selected to meet the needs of the domestic market. Above all, unlike the gas sector, the electricity industry does not rely on foreign production sites. Power transmission routes are fundamentally different from motorways, railways, and gas pipelines. Whilst the mere existence of other railway routes does not have any direct effect on the congestion problem or the problem of internalizing the externality costs on a given route section,25 the situation is totally different in power transmission. Here, the scope of the externality costs cannot be restricted to the direct transmission path between an entry point and an exit point. It rather depends to a large extent on the simultaneous generation (feeding into the system) and withdrawal at the various entry and exit points and on the fixed overall system parameters (voltage restrictions, etc.) in the network. This can be illustrated by a simple example (Hogan, 1992, p. 217): in an interconnected power grid, it is not possible to transport power directly from a particular entry point to a particular exit point. The current will rather choose a path through the grid, which offers the lowest resistance. At least part of the power will therefore not choose the shortest connection (contract path). So which route (or detours) the current will actually take does not just depend on the transmission capacities and resistances of the different lines but more so on the feed and withdrawal

24

In order to move the gas over long distances, transmission pipeline systems have compressor stations to keep boosting the gas pressure. 25 Route-based congestion externalities have long been investigated in traffic economies. Participants usually ignore the damage (e.g. longer travel times) caused by another vehicle on a certain route. These are physical externalities which cannot be internalized through market prices. One possible step would be to charge a congestion fee to the amount of the externality costs incurred by all other participants as a result of one extra journey. This ensures that each vehicle bears the total costs of its own journey.

Market Power Regulation versus General Competition Law

63

rates at all of the entry and exit points. In an extreme case, the capacity of a transmission line between a particular entry point and an exit point will exclusively be used for power transmission from other entry points (cf. Hogan, 1992, p. 213).26 Transmission and distribution of electricity break down into different voltage levels. Power transmission is restricted to extra high-voltage networks covering what are usually large geographic areas. Its main purpose is to interconnect different power plants to allow utilizing the systems’ economies of scale. In addition, the extra high-voltage networks connect power generators with the local utilities’ distribution grids. 2.5.3.2. Power transmission networks as monopolistic bottlenecks Even after the full opening of the electricity market, power transmission networks continue as horizontally linked natural monopolies. At the transmission level, the degree of bundling (interlinking) is very high. The grids operated by the different power utilities are restricted to particular geographic regions and do not overlap. Integrated utilities operate what is in some cases a very complex transmission network within their own service areas. Inside these service areas there are no competing transmission lines owned by other network operators. Owing to the system advantages of an integrated power transmission network, which are based on the interlinking of all points of generation and consumption within the service area, it is most cost-efficient for a particular geographic area to be supplied by only one utility. So there is no incentive, either from a general economic or from a private investor’s point of view, to build individual parallel lines or networks within a given service area. It is therefore highly unlikely that newcomers will build alternative lines or parallel (partial) networks, even after the full opening of the electricity networks. Moreover, the development of transmission networks involves irreversible costs tied geographically to a particular location. These would be lost if a player were to withdraw from the market, because it would be impossible to recoup the capital if a network were to shut down. The transmission network allows the transport of electricity power at high-voltage over long distances and is considered a natural monopoly with highly sunk investment (Brunekreeft, 2003, p. 232). Accordingly, the power transmission network within a utility’s service area has the features of a monopolistic bottleneck. 2.5.3.3. Active network competition in supra-regional transmission gas pipelines The national supra-regional high-pressure natural gas pipeline networks are imbedded in the international pipeline transmission system. The pipeline networks are backbones linking the international network with the regional and/or local service areas. They also feed into downstream regional grids or supply local distribution system operators and industrial users. Depending on the location of the transmission pipeline, this can either be through direct branch lines or through more or less interconnected regional networks. If it can be shown that regional network operators and/or local distribution companies can choose between at least two different operators of supra-regional gas transmission networks, then it is no longer absolutely necessary to have access to the pipelines of a particular supra-regional gas transmission company, which in turn means there is no bottleneck situation at the gas transmission level.

26

The “loop flow” phenomenon, which is based on Kirchhoff’s laws, is thus essentially the same as the economic problem of system network externality.

64

Electricity Market Reform Table 2.5. Electricity networks (transmission and distribution networks) as monopolistic bottleneck facilities.

Generation Transmission networks Regional/local distribution networks Retail

Natural monopoly

Irreversible costs

— X X

X X X

—

—

It is obvious that competing pipeline backbones can only evolve if freedom to build pipelines is guaranteed by law. If instead, free entry is not allowed due to a restrictive licences policy, which guarantees nationwide monopolies, competing pipeline backbones cannot occur, although the characteristics of monopolistic bottlenecks may not be fulfilled. The competition potential created by project companies and pipeline ownership in undivided shares relies on identical transportation routes. Access to alternative supra-regional transmission networks operated by different companies does not necessarily mean that they provide identical transmission routes. Far more important is the question whether there is access to competing pipeline backbones. A transmission company’s backbone system comprises pipelines with utilization rights under joint project company ownership, pipelines co-owned in undivided shares and pipelines owned by only one company. For these pipelines, the concept of detour traffic is of lesser importance as long as existing alternatives for pipeline access do not generate prohibitively high costs. For operators of regional and local distribution networks, the relevant issue is whether there are several alternatives of access into supra-regional transmission pipeline systems. In contrast to electricity transmission grids, supra-regional gas pipelines do not necessarily fulfil the characteristics of natural monopolies and therefore do not necessarily possess the characteristics of a monopolistic bottleneck. For example, the national supra-regional transmission networks in Germany do not possess the characteristics of monopolistic bottlenecks (Knieps, 2002b). 2.5.3.4. Monopolistic bottlenecks in the regional/local distribution networks Electricity distribution networks deliver at low voltage to the end user. Local distribution networks of gas are highly integrated grids serving gas supply purposes in local service areas. Like other local grids, the routes usually follow the network of roads. In towns and cities, the local distribution mains of gas or electricity are usually installed directly under or alongside the streets. In terms of its geometry, the local grid therefore assumes the same structure as the actual street system. The distribution networks of gas or electricity are therefore considered as regional/local natural monopolies with sunk costs. The results of this section can be summarized in Tables 2.5 and 2.6. 2.6. Disaggregated Regulation of Access to Monopolistic Bottlenecks Insofar as there are monopolistic bottleneck areas in network sectors, they require specific residual regulation in order to discipline the remaining market power. In particular, symmetric access to monopolistic bottleneck areas for all active and potential suppliers of network services must be guaranteed, so that competition can become fully effective on all complementary markets.

Market Power Regulation versus General Competition Law

65

Table 2.6. Gas distribution as monopolistic bottlenecks.

Extraction (production) Supra-regional transmission pipelines Regional and local distribution Retail

Natural monopoly

Irreversible costs

— X/—

X X

X —

X —

In contrast to networks under competition, the market power involved in monopolistic bottlenecks fundamentally disturbs the bargaining processes on access conditions. One extreme alternative could be (vertical) foreclosure of competitors on a complementary service market. Such a tying can be used as a method of price discrimination, enabling a monopolist to earn higher profits (Posner, 1976; Chapter 8). Another way of abusing market power in the bargaining process on access conditions is to provide insufficient network access quality or excessive interconnection charges. An example of inferior access conditions is lower quality access to railroad tracks. Monopolistic access charges are another danger when market power due to monopolistic bottlenecks is involved. 2.6.1. Application of regulatory instruments to the monopolistic bottleneck basis The effect of a total refusal of access to monopolistic bottleneck facilities can also be achieved by providing access only at prohibitively high tariffs. This shows that an effective application of the essential facilities doctrine must be combined with a suitable regulation of access conditions to bottlenecks with regard to price, technical quality, and timeframe. However, the fundamental principle of such a regulatory policy should be to strictly limit regulatory measures to those network areas where market power potential does indeed exist. A regulation of access tariffs to monopolistic bottlenecks must therefore not lead to a regulation of tariffs in network areas without market power potential. There are two further issues that have to be taken into account. On the one hand, the existence of competition on the service level should not lead to the conclusion that there is no market power potential on the upstream network level, as long as the latter fulfils the criteria of a monopolistic bottleneck (cf. Brunekreeft, 2003, pp. 89f.). On the other hand, there is the question of the minimum regulatory depth necessary to guarantee non-discriminatory access to essential facilities, without, however, disproportionately interfering with the property rights of the regulated firm.27 2.6.2. Price-level regulation of access charges The identification of monopolistic bottlenecks is always based on an intramodal perspective, a decisive factor being the need for complementary service providers to have nondiscriminatory access to such facilities. However, the existence of monopolistic bottleneck 27

Basically one has to differentiate between, on the one hand, the question whether, due to a monopolistic bottleneck, network-specific market power exists, and, on the other hand, the question what kind of regulatory intervention is suitable. Thus the so-called Hausman–Sidak test argues that a regulatory obligation to unbundle the local loop is not justified, if, even without unbundling, the incumbent is not able to exercise market power with regard to providing telecommunications services to end users (cf. Hausman and Sidak, 1999, pp. 425f.; Hausman, 2002, p. 138).

66

Electricity Market Reform

facilities does not necessarily guarantee that there will be long-term economic profits. Firstly, there is the possibility of the “necessary case”, where even unregulated infrastructure providers are unable to meet their costs. Secondly, competition between modes can severely limit an infrastructure provider’s profit potential. The reference point for regulatory rules concerning access charges should be the coverage of the full costs of the monopolistic bottleneck (in order to guarantee the viability of the facility). In particular, when alternatives to bypass essential facilities are absent, the costcovering constraint may not be sufficient to forestall excessive profits. Therefore, the instrument of price-cap regulation should be introduced (cf. e.g. Beesley and Littlechild, 1989). Its major purpose is to regulate the level of prices, taking into account the inflation rate (consumer price index) minus a percentage for expected productivity increase. It seems important to restrict such price-cap regulation to the bottleneck components of networks, where market power due to monopolistic bottlenecks is really creating a regulatory problem. In other subparts of networks price setting should be left to the competitive markets. Regulation of infrastructure access charges should be limited exclusively to price capping. The basic principle underlying price-capping regulation is that price levels should be regulated in areas where there is network-specific market power. The benefits of price capping in terms of efficiency improvements and future investment activities can only unfold if price capping is applied in its “unadulterated” form and not combined with input-based profit regulation. Individual pricing agreements amount to over-regulation that is harmful to competition. 2.6.3. Flexible pricing structures for network access The question remains whether regulators should also be allowed to prescribe pricing rules focussing on tariff structures within monopolistic bottlenecks. There are serious arguments for regulators to refrain from detailed tariff regulation. In the first place, firms should have the flexibility to design (Pareto superior) optional tariff schemes (e.g. Willig, 1978). Pricing rules prescribed by the regulator could induce inefficient bypass activities. For example, a first pricing rule could be access tariffs according to the long-run average costs of the essential facility. Since in such a case a differentiation among different user groups according to different price elasticities is not possible, incentives for inefficient bypass of the bottleneck facility may be created for certain user groups. A second pricing rule would be access pricing according to the Ramsey pricing principle. Mark-ups on the marginal costs of access to the monopolistic bottlenecks are chosen according to the elasticity of demand for network access in order to maximize social welfare given the cost-covering constraint. However, Ramsey prices could become unsustainable, even if applied strictly to monopolistic bottlenecks. The technological trend toward the unbundling of monopolistic bottleneck components increases the possibilities for inefficient bypass. Secondly, the danger arises that regulators extend the regulatory basis to include the contestable subparts of networks. From the point of view of increasing static (short run) efficiency such behavior could even be justified by welfare theory. It is well known that efficiency distortions caused by applying Ramsey pricing can be reduced by extending the regulatory basis (e.g. Laffont and Tirole, 1994). Nevertheless, such an endeavor would in fact mean a return to fully regulated networks, including priceand entry-regulation of the contestable subparts. As such, this would not be a suitable response to deregulation (e.g. Damus, 1984). Regulatory authorities should not force firms to apply specific pricing rules, such as Ramsey prices or two-part tariffs, as this would hamper their quest for innovative pricing systems. It is always possible that better rules will be found in future. The design of pricing rules should be part of the decision-making process of the firms.

Market Power Regulation versus General Competition Law

67

2.6.4. End-to-end regulation versus disaggregated regulation Regulatory instruments can be differentiated according to whether they are limited to the bottleneck areas (disaggregated regulation) or applied globally (end-to-end), including the competitive segments (e.g. Laffont and Tirole, 2000; Chapter 4). Since the application of regulatory rules is not costless and may also be abused strategically to disturb market forces, the advantage of the disaggregated regulatory approach is the strict limitation of the regulatory basis to bottleneck services. Its disadvantage, however, is that incentives may be created to discriminate against firms in vertically related competitive segments (e.g. Mandy, 2000). This should be kept in mind when designing adequate rules for disaggregated bottleneck regulation. From an economic policy point of view, the use of ex ante sector-specific regulation involves massive interference with the market process and must therefore be supported by a well-founded justification. Even if the nature of the network means that the bottleneck areas are complementary to the other parts of the network, there is no reason whatsoever for end-to-end regulation and a general use of regulatory tools. Both the findings of network economics and the experience with different network sectors show that tailor-made bottleneck regulation is the only way. Generally, a distinction has to be made between the existence of network-specific market power due to monopolistic bottlenecks and the question as to whether this market power is transferred to complementary parts of the market. Even if a transfer of market power from a bottleneck to other partial markets were incentive compatible, this would certainly not mean that the bottleneck and the other partial markets belong to the same market. The basic idea of the disaggregated regulatory approach employed in network economics is the very fact that it is possible to distinguish between those parts of the network that constitute bottlenecks and those parts that are characterized by active and potential competition. The all-important task then is to ensure adequate regulation of bottlenecks in order to enable equal opportunities for competition on the other markets. It is well known from the positive theory of regulation that regulators have strong incentives to over-regulate, mix regulatory instruments in an unsuitable way, favor the application of detailed regulation and call for a heavy-handed supervision of firms (e.g. Stigler, 1971; Knieps, 1998). This is the very reason why an a priori “framing” decision to limit the regulatory basis to some extent is of particular importance. This leads to the disaggregated regulatory approach, which not only identifies networkspecific market power properly as monopolistic bottlenecks, but also designs a combination of regulatory instruments limited to the bottleneck (Knieps, 1997a, p. 331). Price-cap regulation limited to monopolistic bottleneck services (wholesale services) combined with accounting separation and technical regulation is sufficient to deal with the problem of non-discriminatory access. Although access regulation cannot be perfect, it moves regulatory attention into the right direction. The aim of future regulatory policy should not be the global regulation of markets. Instead, only a disaggregated regulation of monopolistic bottlenecks is justified. The aim is then to localize the market power in monopolistic bottleneck areas and discipline this market power by regulatory intervention. Asymmetry of market power due to monopolistic bottleneck facilities, however, does not by itself require asymmetric regulation. Instead, the symmetry principle requires that all firms have access to capacities of monopolistic bottlenecks on terms identical to those of the incumbent (non-discriminatory access). The symmetry principle demands that only bottleneck facilities are regulated, irrespective of whether the owner is the incumbent or a newcomer (Shankerman, 1996, p. 5).

68

Electricity Market Reform

2.6.5. A critical appraisal of access holidays 2.6.5.1. The avoidance of over-regulation In recent years the focus of regulatory attention has increasingly shifted toward the incentives for investment. From an economic point of view the relation between the pricing of access to monopolistic bottlenecks and its linkage with investment incentives has to be analyzed (e.g. Newbery, 2000; Valletti, 2003). In the European Union (EU) review on telecommunications the access directive of March 2002 indicates the necessity that: “National regulatory authorities shall take into account the investment made by the operator and allow him a reasonable rate of return on adequate capital employed, taking into account the risks involved”.28 The new EU directives on the regulation of electricity and gas networks also focus on transmission and distribution tariffs allowing the necessary investments to ensure the viability of networks.29 Access holidays mean a significant period during which an investor is free from access regulation. The idea is that such holidays will increase investment incentives by allowing profits unhindered by regulatory intervention (Gans and King, 2003, p. 164). Access holidays can only be a relevant concept, if regulatory problems of network-specific market power still exist. With respect to market power two questions have to be considered. Firstly, does a new investment create network-specific market power? If not – as for example in the area of LNG terminal projects – sector-specific regulation is superfluous. From this perspective, Art. 22 of the Directive 2003/55/EC on the exemptions from access regulation seems convincing, because it requires that the exemption is not detrimental to competition. The efficient operation of exemption from regulated third party access, therefore, depends on the regulator’s ability to distinguish between essential facilities and projects subject to competition (Hernández and Gandolfi, 2005, p. 6). Secondly, do new investments phase out the bottleneck nature of the existing network infrastructure? An important example comes from the telecommunications sector. Since the comprehensive opening of the telecommunications market, the pressure of innovation has increased in local networks, too. This has led to considerable variety in technological platforms, for example, optical fibre, wireless networks, CATV networks, satellite technology, and an increase in the variety of network access products. Due to these rapid developments the local loop facilities in bigger cities and agglomerations are increasingly loosing their character of monopolistic bottlenecks. Although it is not possible at this point to predict exactly how long it will take for the monopolistic bottlenecks in the local loop to disappear completely, there cannot be any doubt that the regulation of monopolistic bottlenecks has to be viewed in a dynamic context, so that the potential for phasing out sector-specific regulation in telecommunications can be fully exhausted (e.g. Knieps, 2004). 2.6.5.2. Are access holidays justified? The basic argument in favor to access holidays is the conjectured failure of regulatory contracts. Due to the sequential nature of investment decisions (ex ante) and regulation of 28

Directive 2002/19/EC of the European Parliament and of the Council on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive), OJ L108/7, 24.4. 2002. 29 Directive 2003/54/EC of the European Parliament and of the Council concerning common rules for the internal market in electricity and repealing Directive 96/92/EC, OJ L 176/37, 15.7.2003, Art. 23/2 (a). Directive 2003/55/EC of the European Parliament and of the Council concerning common rules for the internal market in natural gas and repealing Directive 98/30/EC, OJ L 176/57, 15.7.2003, Art. 25/2 (a).

Market Power Regulation versus General Competition Law

69

access tariffs (ex post), a regulation-induced hold-up problem would arise. The truncation problem would result to reward only ex post-successful projects, whereas the ex ante risks of project failure would not be compensated. The question arises whether access holidays are the adequate answer to this problem. Due to the past dependency of network infrastructures the hypothetical scenario of greenfield approaches of new infrastructure networks overstates the ex ante risks, although the incentive problem for the gradual renewal of bottleneck infrastructures should not be ignored. From an investor’s point of view all relevant ex ante risks should be compensated, including the option value of waiting to invest. The design of credible regulatory contracts focussing on the financial viability of the networks is required (e.g. Knieps, 2005). An important cornerstone of the EU directives is the financial viability of the networks.30 Access tariffs in the electricity and gas networks should allow the necessary investments in the networks.31 In telecommunications national regulatory authorities are obliged to take into account the investment made by the operator and allow him a reasonable rate of return on adequate capital employed, taking into account the risk involved (Access Directive 2002/19/EC, Art. 13(1)). For railroads, full recovery of the infrastructure costs can be accomplished via access charges only, without state funding. Where a contractual agreement for state funding exists, the terms of the contract and the structure of the payment have to be agreed upon in advance to cover the whole of the contract period (Directive 2001/ 14/EC, Art. 6/1, Art. 6/4, Art. 8/1). It can be shown that the problem of regulatory opportunism is not caused by the nature of ex ante irreversible investments per se, but is based on the more general problem that regulatory agencies cannot be committed to welfare-maximizing behavior. Therefore, the regulatory agencies have to be constrained by statutes to allow the compensation of ex ante risks of irreversible investments. To conclude, the instrument of access holidays becomes superfluous. 2.7. Conclusions In order to analyze the role of sector-specific regulation compared to general competition policy in the liberalized energy markets a disaggregated approach is developed. The theory of monopolistic bottlenecks constitutes the theoretical reference point for this analysis in order to identify stable network-specific market power. A survey is devoted to of the localization of monopolistic bottlenecks in different network industries. Finally, the concept of disaggregated price-cap regulation of monopolistic bottlenecks is provided avoiding end-to-end regulation as well as access holidays. Appendix 2.A1. Empirical/Econometric Approaches The traditional approaches to proving with sufficient certainty the existence of stable market power by means of empirical/econometric methods for a certain industry have, to 30

For railway infrastructures see Directive 2001/14/EC, Art. 8; for electricity grids see Directive 2003/54/EC, Art. 19; for natural gas pipelines see Directive 2003/55/EC, Art. 18; for telecommunications networks see Access Directive 2002/19/EC, Art. 10; for slot allocation at Community airports see Council-Regulation (EEC) No. 95/93. 31 Directive 2003/54/EC, Art. 22/2a; Directive 2003/55/EC, Art. 25/2a.

70

Electricity Market Reform

date, not developed far enough to serve as a basis for justifying regulatory intervention. Also to be considered in this context is the usual point of reference in a market economy that in markets the 1st order error (“false positive”), that is, the fact that the authorities interfere with the competition process by way of regulatory action even though competition is working and there is no need whatsoever for state action, is particularly grave (Knieps, 1997a, p. 51). The following quotation characterizes the concept of market power established in antitrust literature: “The term ‘Market Power’ refers to the ability of a firm (or a group of firms, acting jointly) to raise price above the competitive level without losing so many sales so rapidly that the price increase is unprofitable and must be rescinded” (Landes and Posner, 1981, p. 937). Stable market power is characterized by the fact that the long-term characteristics of the market under review (particularly the production and demand conditions) allow stable profits without these being competed away (e.g. by arbitrage activities). Quite apart from this are short-term profits which can occur as a result of short-term characteristics of the market under review and are then rapidly competed away by other suppliers. These kind of unstable or fleeting profits are, however, very difficult to differentiate from pioneer profits. The much used but still problematic method of “proving” market power in antitrust cases consists in first defining the relevant market in which the market share of the company accused is to be determined; second, calculating this market share; and third, deciding whether this market share is sufficiently large to allow conclusions to be drawn regarding the necessary extent of market power. However, this method is unsuited as a basis for deriving stable market power and for establishing a regulatory policy in network industries. Market shares are not a reliable criterion for market power.32 This can already be explained by way of the known phenomenon of reversed causality. Instead of being able to impose higher prices because of a high market share, low prices lead to a high market share. In the case of free market access, competitors will enter the market at lower prices with the result that a seller with high prices is undercut and loses market shares. Obviously, it would be incorrect to assume market power without hesitation merely because of a high market share (cf. Landes and Posner, 1981, p. 977). In network industries where economies of scale are frequent, the resulting decrease in average costs augments this effect. Nevertheless, there is the risk that the authorities may use the market share criterion to conclude automatically the existence of a dominant market position. Since the size of the market share cannot be used conclusively to infer the existence of stable market power, there is the question of how far econometric models are capable of assessing whether or to what extent stable market power actually exists in an industry. The methods best known and most often used are the structural models and the “reduced form” methods,33 all of which are in themselves consistent and mathematically correct.34 However, stable market power cannot reliably be localized with any of these model approaches. Moreover, applying 32

The fundamental criticism of the structure–conduct–performance approach of traditional industrial economics is summarized extensively (e.g. in Schmalensee, 1989). 33 A critical assessment of these approaches along with extensive literature on the subject is given in Hyde and Perloff (1995). 34 In a structural model, the econometrician assesses all equations of that model. Ideally one would assess a complete model with separate behavioral equations for each company of the industry. If only aggregated industry level data are available, a demand equation, an aggregated cost equation and a balance condition are estimated. A market power assessment based on these methods is extremely sensitive to small changes in the specification of the structural model equations on which they are based. Structural

Market Power Regulation versus General Competition Law

71

these different methods to the same industries typically leads to contradictory results, with at least one or even several tests rejecting the competition hypothesis for most of the industries.35 It is obvious that regulatory intervention based on these results of industry-economic research would lead to interventionism, which would be damaging to the competition process. 2.A2. Game Theory Approaches Another methodological approach aimed at unveiling stable market power in industries leads to game theory. Over the last two decades, game theory models have made significant inroads into industrial economics and competition theory-related research. Yet the extensive literature on game theory models (e.g. Tirole, 1989) fails to provide a basis for reliably localizing stable market power in industries. The result is not surprising if one looks at the strengths and weaknesses of the game theory analyzes.36 On the one hand, there are typically a multitude of alternatives for developing a game theory model, all of which appear equally appropriate a priori. On the other hand, there is typically a multitude of solutions within a certain model approach. However, the advantage of having a multitude of possibilities for modeling interaction between alternative suppliers based on game theory proves problematic when it comes to searching for robust solutions. If market power can be localized for a very specific model specification, but then disappears if there is a minor change to any of the model parameters, then this is no reliable basis for ex ante regulatory intervention. The search for robust characteristics of game theory models, which can also be examined empirically, has only just begun (cf. Sutton, 1990). Given the present state of research, it is unsuited as a theoretical basis for localizing stable market power and establishing a regulatory framework.

References Areeda, P. and Hovenkamp, H. (1988). “Essential facility” doctrine? Applications, Antitrust Law, 202.3 (Suppl. 1988), 675–701. Bailey, E.E. (1981). Contestability and the design of regulatory and antitrust policy. American Economic Review, 71(2), 178–183.

(continued) models require more extensive data and more explicit assumptions than the “reduced form” methods. A known “reduced form” method was developed by Panzar and Rosse (1987), whose market power test only requires the assessment of a single equation. The Panzar–Rosse method is easier to apply than the structural model approach. The problem with this method, however, is that the correct “reduced form” revenue function is extremely complicated, not linear, and therefore difficult to assess. In the end, a reliable distinction between collusion and competition is not possible. Another known “reduced form” method was developed by Hall (1988). Hall’s method uses results of comparative statistics to test market power, the zero hypothesis being the competition. The central weakness of the Hall approach is that it can only be used to measure the market power of an industry if one can reliably assume constant returns to scale. The reliability of the results is extremely sensitive to deviations in the constant returns to scale. Both methods differ both with regard to the scope of the data requirements and in the underlying assumptions. 35 “For poultry, butter, cheese, we cannot reject competition based on any of the tests shown in Table IV. For all other industries, one or more tests reject competition or are implausible (Hall estimate of ␮ for red meat and flour). For these other industries, there is little consistency in results across the methods, though, in most cases, the results are consistent with some form or another of oligopoly or monopolistic competition” (Hyde and Perloff, 1995, p. 481). 36 For an illustrative overview, see Fisher (1989).

72

Electricity Market Reform

Bailey, E.E. and Baumol, W.J. (1984). Deregulation and the theory of contestable markets. Yale Journal on Regulation, 1, 111–137. Bailey, E.E. and Panzar, J.C. (1981). The contestability of airline markets during the transition to deregulation. Law and Contemporary Problems, 44, 125–145. Bailey, E.E. and Williams, J.R. (1988). Sources of economic rent in the deregulated airline industry. Journal of Law and Economics, 31, 173–202. Bain, J.S. (1956). Barriers to New Competition. Harvard University Press, Cambridge, MA. Baumol, W.J. (1977). On the proper cost test for natural monopolies in a multiproduct industry. American Economic Review, 67, 809–822. Baumol, W.J. (1982). Contestable markets: an uprising in the theory of industry structure. American Economic Review, 72, 1–15. Baumol, W.J. (1996). Predation and the logic of the average variable cost test. Journal of Law and Economics, 39, 49–72. Baumol, W.J. and Willig, R.D. (1986). Contestability: developments since the book. Oxford Economic Papers, Special Supplement, November, 9–36. Baumol, W.J. and Willig R.D. (1999). Competitive rail regulation rules, should price ceilings constrain final products or inputs? Journal of Transportation Economics and Policy, 33(1), 43–54. Baumol, W.J., Panzar, J.C. and Willig, R.D. (1982). Contestable Markets and the Theory of Industry Structure. Harcourt Brace Jovanovich, San Diego. Baumol, W.J., Panzar, J.C. and Willig R.D. (1983). Contestable markets: an uprising in the theory of industry structure: reply. American Economic Review, 73(3), 491–496. Beesley, M.E. and Littlechild, S.C. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20, 454–472. Brunekreeft, G. (2003). Regulation and Competition Policy in the Electricity Market – Economic Analysis and German Experience, Nomos, Baden-Baden. Carlton, D.W. (2004). Why barriers to entry are barriers to understanding. American Economic Review, 94, 466–470. Caves, D.W., Christensen, L.R. and Tretheway, M.W. (1984). Economies of density versus economies of scale: why trunk and local airline costs differ. Rand Journal of Economics, 15(4), 471–489. Chadwick, E. (1859). Results of different principles of legislation and administration in Europe; of competition for the field, as compared with competition within the field, of service. Journal of the Royal Statistical Society, 22, 381–420. Damus, S. (1984). Ramsey pricing by U.S. railroads – can it exist? Journal of Transport Economics and Policy, 18, 51–61. Davies, J.E. (1986). Competition, contestability and the liner shipping industry. Journal of Transport Economics and Policy, September, 299–312. Debreu, G. (1959). Theory of Value: An Axiomatic Analysis of Economic Equilibrium. Yale University Press, New Haven and London. Demsetz, H. (1968). Why regulate utilities? Journal of Law and Economics, 11, 55–65. Fisher, F.M. (1989). Games economists play: a non-cooperative view. Rand Journal of Economics, 20(1), 113–124. Gans, J. and King, S. (2003). Access holidays for network infrastructure investment. Agenda, 10(2), 163–178. Graham, D.R., Kaplan, D.P. and Sibley, D.S. (1983). Efficiency and competition in the airline industry. Bell Journal of Economics, 14, 118–138. Hall, R.E. (1988). The relationship between price and marginal cost in U.S. industry. Journal of Political Economy, 96, 921–947. Hausman, J. (2002). Internet-related services: the results of asymmetric regulation. In: R.W. Crandall and J.H. Alleman (Eds.), Broadband – Should We Regulate High-Speed Internet Access? Brookings Institution Press, Washington DC, pp. 129–156. Hausman, J. and Sidak, J.G. (1999). A consumer-welfare approach to the mandatory unbundling of telecommunications networks. Yale Law Journal, 109, 417–505. Hernández, F. and Gandolfi, M. (2005). EU exemptions to TPA for new gas infrastructures. Energy Regulation Insights, July(24), NERA, London.

Market Power Regulation versus General Competition Law

73

Hogan, W.W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4, 211–242. Hyde, C.E. and Perloff, J.M. (1995). Can market power be estimated? Review of Industrial Organization, 10, 465–485. Joskow, P.L. and Klevorick, A.K. (1979). A framework for analyzing predatory pricing policy. Yale Law Journal, 89, 213–270. Knieps, G. (1997a). Phasing out sector-specific regulation in competitive telecommunications. Kyklos, 50(3), 325–339. Knieps, G. (1997b). The concept of open network provision in large technical Systems. EURAS Yearbook of Standardization, 1, 357–369. Knieps, G. (1998). Costing and pricing of interconnection services in a liberalized european telecommunications market. In: Telecommunications Reform in Germany: Lessons and Priorities. American Institute for Contemporary German Studies, Washington DC, pp. 51–73. Knieps, G. (2002a). Does the system of letter conveyance constitute a bottleneck resource? Discussion Paper No. 101, Institute of Transport Networks and Regional Policy, University of Freiburg, Germany. Knieps, G. (2002b). Wettbewerb auf den Ferntransportnetzen der deutschen Gaswirtschaft – Eine netzökonomische Analyse. Zeitschrift für Energiewirtschaft (ZfE), 26(3), 171–180. Knieps, G. (2004a). Information and communication technologies in Germany: is there a remaining role for sector-specific regulations? In: A. Moerke and C. Storz (eds.). Institutions and Learning in New Industries, RoutledgeCurzon, 2006 (forthcoming). Knieps, G. (2005). Telecommunications markets in the stranglehold of EU regulation: on the need for a disaggregated regulatory contract. Journal of Network Industries, 6(2), 75–93. Knieps, G. (2006). Delimiting Regulatory Needs. In: OECD/ECMT Round Table 129, Transport Services: The limits of the (De) regulation, OECD Publication service. Paris, pp. 7–31. Knieps, G. and Brunekreeft, G. (eds.) (2003). Zwischen Regulierung und Wettbewerb: Netzsektoren in Deutschland, 2. Edition. Physica-Verlag, Heidelberg. Knieps, G. and Vogelsang, I. (1982). The sustainability concept under alternative behavioral assumptions. Bell Journal of Economics, 13(1), 234–241. Kuhlmann, A. and Vogelsang, I. (2005). The German electricity sector – finally on the move? CESifo, DICE Report, 3(2), 30–39. Laffont, J.-J. and Tirole, J. (1994). Access pricing and competition. European Economic Review, 38, 1673–1710. Laffont, J.-J. and Tirole, J. (2000). Competition in Telecommunications. The MIT Press Cambridge, Massachusetts, London. Landes, W.M. and Posner, R.A. (1981). Market power in antitrust cases. Harvard Law Review, 94, March, 937–997. Lipsky, A.B. and Sidak, J.G. (1999). Essential facilities. Stanford Law Review, 51, 1187–1249. Mandy, D.M. (2000). Killing the goose that may have laid to the golden egg: only the data know whether sabotage pays. Journal of Regulatory Economics, 17(2), 157–172. McAfee, R.P., Mialon, H.M. and Williams, M.A. (2004). What is a barrier to entry? American Economic Review, 94, 461–465. Morrison, S.A. and Winston, C. (1987). Empirical implications and tests of the contestability hypothesis. Journal of Law and Economics, 30, 53–66. Newbery, D.M. (2000). Privatization, Restructuring, and Regulation of Network Utilities. Cambridge (MA), London. Panzar, J.C. and Rosse, J.N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35, 443–456. Panzar, J.C. and Willig, R.D. (1977). Free entry and the sustainability of natural monopoly. Bell Journal of Economics, 8, 1–22. Posner, R.A. (1976). Antitrust Law: An Economic Perspective. University of Chicago Press, Chicago. Schmalensee, R. (1989). Inter-industry studies of structure and performance. In: R. Schmalensee and R. Willig (eds.), Handbook of Industrial Organization. North-Holland, Amsterdam, pp. 951–1009. Schmalensee, R. (2004). Sunk costs and antitrust barriers to entry. American Economic Review, 94, 471–475.

74

Electricity Market Reform

Schwartz, M. (1986). The nature and scope of contestable theory. Oxford Economic Papers, Special Supplement, November, 37–57. Schwartz, M. and Reynolds, R.J. (1983). Contestable markets: an uprising in the theory of industry structure: comment. American Economic Review, 73(3), 488–490. Shankerman, M. (1996). Symmetric regulation for competitive telecommunications. Information Economics and Policy, 8, 3–23. Stigler, G.J. (1968). Barriers to entry, economies of scale, and firm size. In: G.J. Stigler. The Organization of Industry. Irwin, Homewood, IL, pp. 67–70. Stigler, G.J. (1971). The theory of economic regulation. Bell Journal of Economics, 2, 3–21. Stiglitz, J.E. (1987). Technological change, sunk costs and competition. Brookings Papers on Economic Activity, 3, 883–947. Sutton, J. (1990). Explaining everything, explaining nothing? Game theoretical models in industrial economics. European Economic Review, 34, 505–512. Tirole, J. (1989). The Theory of Industrial Organization, 2nd printing. MIT Press, Cambridge. Valletti, T.M. (2003). The theory of access pricing and its linkage with investment incentives. Telecommunications Policy, 27, 659–675. Weitzman, M.L. (1983). Contestable markets: an uprising in the theory of industry structure: comment. American Economics Review, 73, 486–487. Weizsäcker, C.C. von (1980a). A welfare analysis of barriers to entry. Bell Journal of Economics, 11, 399–420. Weizsäcker, C.C. von (1980b). Barriers to Entry: A Theoretical Treatment, Springer, Berlin. Willig, R.D. (1978). Pareto superior nonlinear outlay schedules. Bell Journal of Economics, 9, 56–69. Willig, R.D. (1980). What can markets control? In: R. Sherman (Ed.), Perspectives on Postal Service Issues. American Enterprise Institute for Public Policy Research, Washington, pp. 137–159.

PART II Trailblazers

This page intentionally left blank

Chapter 3 Chile: Where It All Started RICARDO RAINERI Departamento Ingeniería Industrial y de Sistemas, Escuela de Ingeniería, Pontificia Universidad Católica de Chile, Santiago, Chile

Chapter summary This chapter describes the pioneering deregulation and privatization process of the Chilean electric industry, which began in the 1980s. Over the years, Chile has had its share of success and failures, and offers an early example of the difficulties in dealing with three distinct segments of the value chain, namely generation, transmission, and distribution, along with a system operator. In the Chilean context, power generation was left to competitive investors, transmission was turned into an open access regime, and distribution was left as a regulated natural monopoly. The system was put to a test during a severe drought in the late 1990s and subsequent crises, including chronic shortages of natural gas from neighboring Argentina. After 25 years of fine-tuning, the sector still goes under continuous adjustments which respond to a learning process based on what have worked in the past and what have not, and thus it offers useful insights for policy makers and regulators around the world.

3.1. Introduction Chile is generally credited as the place where electricity supply industry (ESI) market reform started. With the enactment of DFL No. 11 in 1982, a new institutional framework for a decentralized and privately owned ESI was introduced. DFL No. 1 recognized three distinct segments: generation, transmission, and distribution. Power generation was considered as a competitive business, transmission was to be governed as an open access regime allowing all generators a non-discriminatory use of available transmission capacity, and distribution was left to be regulated as a natural monopoly. To coordinate the operations of competitive generators in an open access transmission network, an independent system operator – in this case Centro de Despacho Económico de Carga (Economic Load Dispatch Center, CDEC) – was created. These were considered radical ideas at the time, but now a standard practice.

1

General Law of Electric Services, Decree-Law No. 1 of 1982 from the Ministry of Mines (DFL No. 1).

77

78

Electricity Market Reform

The law defines three markets where generators can sell power and energy. The first market, between generators and distribution companies, is for small consumers, who pay regulated energy and power prices to their distribution company. The second market is for large consumers, who freely agree electricity supply contracts with generators or distribution companies. And a third market in the CDEC is the one where generators trade power and energy to fulfill their electricity supply contracts, in which case the prices of the energy transfer between the companies are decided by the CDEC according to the system’s marginal cost of generation, based on generators’ declared costs, while power transfers are valued according to the National Energy Commission (Comisión Nacional de Energía, CNE) regulated capacity price paid by small consumers in their electricity tariffs. Not only does Chile lead the world in the ESI market reform, but it also was among the first to experience natural and man-made crises with serious consequences. The first one, in 1998–1999, was caused by one of the worst droughts experienced by the country which affected a predominantly hydro-based system. The second, in 1999, resulted from an incomplete regulatory framework which caused a series of blackouts by the lack of coordination among power generators. And the third one, started in 2004 and continuing, resulted from political and economic phenomena in neighboring countries which severely affects a reliable supply of imported natural gas from Argentina, essential for an increasing natural gasdependent ESI in Chile. Other well known electricity crisis in a restructed electricity market is the one that suffered California (see Joskow, 2001a and 2001b). This chapter describes the fundamentals leading to the deregulation and privatization process of the ESI in Chile, the key crises that through these years have tested its normal operation, the policies and politics that have been adopted to address the problems, and the lessons we have learned from the whole process. Having been the first and battered a lot, Chile offers many useful insights for policy makers and regulators around the world. This chapter is organized into four sections that deal with the Chilean ESI. After the Introduction, Section 3.2 introduces the Chilean ESI, Section 3.3 describes the electric crises which affected the Chilean ESI, and Section 3.4 provides the final comments and key lessons which can be learned from the Chilean experience. 3.2. The Chilean ESI Chile is a thin and long country with a 756,950 km2 territory and a population of almost 16 million. In 2004, Chile had a total electricity consumption of 46,114 GWh and an installed generation capacity of 11,561.4 MW. Its ESI consists of four distinct interconnected systems (Fig. 3.1 and Table 3.1) that are isolated from each other: ●

●

Sistema Interconectado del Norte Grande (Greater North Interconnected System, SING) in the north. Sistema Interconectado Central (Central Interconnected System, SIC) covering the central part of the country including the country’s capital city, Santiago.

And two smaller systems in the south: ● ●

Sistema de Aysen (Aysen System). Sistema de Magallanes (Magallanes System).

The SING and the SIC account for the lion’s share of the installed generation and consumption, with the other two systems serving much smaller populations, as shown in Table 3.1.

Chile: Where It All Started

79

Fig. 3.1. ESI systems in chile. Source: Picture elaborated by the author with CNE data.

3.2.1. Key characteristics For the most part, this chapter is focused on the two large systems, the SIC and the SING, which are briefly described below. 3.2.1.1. Sistema Interconectado del Norte Grande The SING is a rather unusual system consisting of a predominantly mining/industrial load, in the order of 90%. It serves a number of large and isolated copper mines, some with individual loads as large as 300 MW. The ESI is mostly based on thermal generation located on the coastal edge, which experienced a significant growth between 1993 and 2004

80

Electricity Market Reform

Table 3.1. Electric systems in Chile (2004).

Installed capacity (MW)

Transmission Peak load lines (km) (MW)

% of population and consumption attributed to small users (residential, Demand commercial and (GWh) industrial)

SING

3596*

4889*

1567

11,240

5.6% of the 99.6% national population, thermal which represents 0.4% hydraulic 10% of the system’s consumption

SIC

7975*

8745

5431

34,602

93% of the national population, which represents 60% of the system’s consumption

41% thermal 59% hydraulic

Sistema de Aysen

34

–

18

82

20,000 customers

44% thermal 50% hydraulic 6% wind

Sistema de Magallanes

65

8.5

33

189

46,000 customers

100% thermal

Type of generation

*April 2005.

(Table 3.2). Today it counts with three distribution companies and three large independent generators. By international standards Chilean ESI market concentration appears high, far above for example to what happen in Argentina (see in this book the chapter by Isaac Dyner et al. “Understanding The Argentinean and colombian Electricity Markets”). The technological innovation in natural gas power plants, that remarkably increased their efficiency during the last two decades, and the increasing availability in the region of natural gas, together with the chances in the late 1990s to import natural gas from Argentina, makes possible the substitution of coal and petroleum power plants for natural gas combined cycle turbines. Given that natural gas power plants have smaller variable costs, they become the base of generation in the SING. The large investments in natural gas power plants implied that the SING installed capacity increased from 1277 MW in 1997 to 3596 MW in 2004, albeit peak load represents only 40% of the installed capacity. Figure 3.2 shows the growth that the installed capacity experienced with respect to the electricity consumption for the SIC and the SING. Although some of the new natural gas power plants enter the market without electricity supply contracts with the large mining companies or distribution companies, in spite of everything they were able to assure enough revenues to finance their investments. This is because the power plants’ dispatch is made according to the lowest variable cost, independently from the generators’ electricity supply contracts, which assures that more efficient power plants can sell their electricity at least at system’s marginal cost.2 2

Generators received energy and a capacity payment. The energy payment is determined by the power plant dispatched with the higher variable costs, and the capacity payment is defined by a function which considers the availability of the power plants and a regulated capacity price.

81

Chile: Where It All Started Table 3.2. SING. Thermal

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Sales (GWh)

Annual growth (%)

Capacity (MW)

Total (MW)

%

3040 3394 3989 4981 5749 6616 8120 8398 8991 9482 10,480 11,240

11.66 17.53 24.87 15.41 15.09 22.72 3.43 7.06 5.46 10.53 7.25

745 799 1157 1160 1277 1476 2637 3041 3441 3633 3641 3596

735 789 1144 1147 1264 1463 2624 3028 3428 3620 3628 3583

98.7 98.7 98.9 98.9 99.0 99.1 99.5 99.6 99.6 99.6 99.6 99.6

%

MW

%

Peak load (MW)

38 50 56 58 58 59

10 10 13 13 13 13 13 13 13 13 13 13

1.3 1.3 1.1 1.1 1.0 0.9 0.5 0.4 0.4 0.4 0.4 0.4

498 611 747 812 1021 1094 1154 1221 1360 1416 1567

Natural gas MW

1004 1519 1919 2112 2112 2112

Hydraulic

Source: CDEC SING, CNE.

3.2.1.2. Sistema Interconectado Central The SIC is the largest electric system in the country with 31 distribution companies, where 60% of its consumption corresponds to small regulated end users. In the period from 1990 to 2004, the SIC electricity demand almost tripled, from 12,512 to 34,602 GWh (Table 3.3 and Fig. 3.2), and installed capacity doubled, from 3195 to 7867 MW, where peak load have fluctuated between 62% and 79% of the installed capacity. Today the SIC have three large independent generators and few smaller companies or co-generators. Until 1996 the SIC depended on hydro generation which represented 75% of installing capacity, implying that electricity supply depends on what happen with hydrologic conditions. This situation changed somehow in 1997 when, motivated by private investors, the country starts importing natural gas from Argentina for the new power plants. With that and up to year 2004, the share of hydro capacity on total capacity decreased to 60%, while the share of natural gas power plants increased from 0% to 23%. Figure 3.2 shows that, contrary to that happened in the SING, the growth in generation capacity in the SIC has been more modest, and sometimes it lagged behind the growth in electricity consumption. Despite the rapid growth of thermal generation, the SIC still remains as a predominantly hydro-based system (Fig. 3.3), during the period 1985–2004 the average contribution of hydro generation to total generation is more than 75%. During unusually wet seasons, this figure could rise to 97% as in 1992, but it could be as low as 48% as in 1999 (Fig. 3.4). The volatility of hydrologic conditions is among the factors that complicate the Chilean ESI – not unlike other hydro-dominated systems such as the Nordic countries, Brazil, and Colombia. 3.2.2. Factors leading to the liberalization of the ESI The period between 1973 and 1990, the military government of Augusto Pinochet, inspired free market policies in Chile, leading to the reform of the ESI. Much of this was driven

82

Electricity Market Reform

550 500

SING capacity and demand growth index (SING 1993 ⫽ 100) SING demand SING capacity

450 400 350 300 250 200 150 100 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 (a) SIC capacity and demand growth index (SIC 1990 ⫽ 100) 280 260

SIC demand SIC capacity

240 220 200 180 160 140 120 100 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 (b) Fig. 3.2. SING and SIC capacity and demand growth indexes. Source: CNE, CDEC SING and CDEC SIC.

by ideologies that promoted competition through the active participation of the private sector in the economy, with the government retaining only a subsidiary role. To understand what happened during this period, one must put the ESI reform in a historic context. The change in the economic policy orientation during the military government reversed the previous 40-year trend of increasing government’s intervention in economic activities. It was during Allende’s Government (1970–1973) that State’s intervention in economic activity reached a maximum, where 100% of public services companies were owned by the government,3 with end users’ tariffs subsidized and the companies’ employment policy highly influenced by the political preference of the worker.

3

See Hachette and Lüders (1994).

83

Chile: Where It All Started Table 3.3. SIC. Thermal

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Sales (GWh)

Annual growth (%)

Capacity Total (MW) (MW)

%

MW

%

MW

%

Peak load (MW)

12,365 12,516 13,811 15,272 16,549 17,672 19,027 20,869 22,434 24,246 25,530 27,654 29,144 30,335 32,076 34,602

1.2 10.35 10.58 8.36 6.79 7.67 9.68 7.50 8.08 5.30 8.32 5.39 4.08 5.74 7.88

2942.1 3195.1 3831.1 3831.1 3889.9 3893.4 4083.6 4858.5 5266.8 6242.4 6695.1 6652.8 6579.2 6737.2 6996.2 7867.4

0.25 0.27 0.22 0.22 0.20 0.19 0.22 0.25 0.30 0.38 0.42 0.39 0.39 0.40 0.42 0.40

– – – – – – – – 379 1119 1119 1359 1359 1467 1467 1847

– – – – – – – – 0.07 0.18 0.17 0.20 0.21 0.22 0.21 0.23

2221 2334 2994 2994 3126 3153 3176 3667 3705 3892 3906 4030 4030 4055 4055 4695

0.75 0.73 0.78 0.78 0.80 0.81 0.78 0.75 0.70 0.62 0.58 0.61 0.61 0.60 0.58 0.60

2270 2273 2375 2632 2819 3070 3235 3497 3773 3991.4 4185.5 4516 4694 4878 5162 5430.8

721 861 837 837 764 741 908 1192 1562 2351 2789 2623 2549 2682 2941 3172

Natural gas

Hydraulic

Source: CDEC SIC, CNE. SIC installed capacity by source 5000 4500

Thermal Hydraulic

4000 3500

MW

3000 2500 2000 1500 1000 500 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fig. 3.3. SIC installed capacity by source. Source: CDEC SIC and CNE.

After the military coup of Augusto Pinochet in September 1973, the Junta focused on reforming the ESI by introducing tariffs that allowed the firms to cover their legitimate costs while encouraging reductions in the number of employees to what can be considered efficient given their production level. At that time, pricing in the ESI was governed by DFL No. 4, enacted in 1959, by which tariffs were based on historical costs. By the end of the 1970s, the

84

Electricity Market Reform SIC generation by source

100.0 9

6

3

5

90.0

7

17

18

13

28

34

80.0

Hydraulic Thermal

14 25 32

40

30

37

41

35 43

52

70.0 60.0 50.0 91

40.0 30.0

94

97

95 84

82

93

87

86 72

66 60

76 68 63

59

71

65 57

48

20.0 10.0 0.0

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 3.4. SIC generation by source. Source: CDEC SIC and CNE.

previous problems were solved to some extent, but some other important problems that needed further changes persisted:4 ● ●

●

●

Still a strong involvement of the State in the development of the industry.5 Increasing monopolization of the industry and its development in one State company, Empresa Nacional de Electricidad S.A. (ENDESA),6 damaging the likelihood to have a competitive industry and implying almost no chances for the government to be an effective counterbalance in the industry. Confusing entrepreneurial and regulatory roles of the State, with private investors at a disadvantage with respect to State companies. Discriminatory tariffs based on historical costs plus a rate of return, lacking from economic, uniform, and transparent procedures.

Until 1978 Chilean electric industry regulation was partially a responsibility of ENDESA; and because this was in conflict with the military government objectives, it created two agencies to take on the regulatory and supervisory roles of the State: ●

4

In 1978 the National Energy Commission (CNE): in-charge of the planning and policy design for the efficient operation of the energy sector and to advise the government on all the aspects of the industry.

See Comisión Nacional de Energía (1989). In 1979, the State participation in the electric sector reached 90% in generation, 100% in transmission and 80% in distribution, demanding more than US$ 200 million of public investments per year to satisfy demand increases. 6 ENDESA was founded by the government in 1943 under the custody of Corporación de Fomento de la Producción (CORFO). CORFO founded in 1939 was originally conceived as a State Development Bank to assist the private sector. But shortly CORFO became the vehicle by which the State becomes involved in the different areas of the economy. CORFO sheltered State companies in the areas of: electricity, telecommunications, chemistry, pulp and forestry, fishing, sugar, coal, computer services, and many others. ENDESA is the result of a national electrification plan developed in 1935 and endorsed in 1936 by the Chilean Engineers Institute. 5

Chile: Where It All Started

●

85

In 1985 the Superintendence of Electricity and Fuels (Superintendencia de Electricidad y Combustibles, SEC): with a supervisory role, and the inspection and security of electric facilities.7

For the definition of regulated prices the responsibility was assigned to the Ministry of Economy, which considers the cost studies of an efficient model company made by CNE. The military government core reform imperatives were the promotion of higher efficiencies through open competition in generation and regulating the distribution tariffs. These ideas have been included in the electric law DFL No. 1 of 1982, and still embody that spirit. Table 3.4 summarizes the major milestones of the Chilean ESI reform since the late 1970s. 3.2.2.1. System operator The final objective of interconnected electric systems is to minimize the global cost of electricity supply by combining loads and generators of different nature, to arbitrage the excess of electricity supply within the different nodes, and sharing the margins of reserve capacity within all the nodes. In Chile, the system operator, CDEC, has been assigned the role to coordinate the operation of interconnected generation and transmission facilities. The SIC (CDEC-SIC) and the SING (CDEC-SING) system operators were created in 1985 and 1993, respectively. The law requires that the concessionaries of interconnected electric systems should coordinate their operation to: ● ● ●

preserve service quality; guarantee the less expensive operation of the system; and guarantee the right way of main transmission and sub-transmission systems established through a concession.8

The governance structure of the CDEC consists of a Board with representatives from generators, transmission companies, and co-generators. This Board is responsible for agreeing the bylaws that complement the law, required to accomplish the objectives of the CDEC. The Board is complemented by an Executive Direction9 which runs the system and determines electricity transfers and prices between the generators. With the existent generation–transmission facilities, the CDEC is responsible for its operation. The different characters of the SIC and the SING – the first with a large hydro capacity by which it can accumulate water from one year to the next, while the second an almost a 100% thermal system – have implied that the tools used for their planning and operation slightly differ but are inspired in the same principles of minimizing the long-term cost of electricity supply, given existent generation–transmission facilities and service reliability requirements. Each system dispatch is made independently of the ownership and the electricity supply contracts of each of the generators. Electricity transfers between generators result from the accounting of the efficient dispatch of the system and the supply contracts of each of the generators, where energy pricing is based on the cost of the less efficient unit being dispatched and capacity pricing is defined by CNE. 7

SEC was created in 1985 with Law No. 18,410 to replace the Superintendence of Electric Services, Gas and Telecommunications. Also, in 1999 with Law No. 19,613, SEC attributions are changed allowing it to impose stricter regulations and sanctions on the electric companies. 8 Law No. 19,940, March 2004, changed this article to “Guarantee the open access of main transmission (backbone) and sub-transmission systems, in compliance with this law”. 9 Divided in an Operations Direction and a Tolls Direction.

86

Electricity Market Reform

Table 3.4. Major events in Chilean electric industry. Event

Year

Comments

Creation of the CNE, Decree-Law No. 2,224

1978

A specialized institution to take on the regulatory and advisory roles on energy issues. The idea is to separate the regulatory, supervisory and entrepreneurial roles of the government.

Nodal prices

1980

First tariff decree to set nodal prices at the level of generation, transmission and distribution.

DFL No. 1

1982

The main reform is embodied in the General Law of Electric Services.

Creation of the Superintendence of Electricity and Fuels (SEC), Law No. 18,410

1985

Replaces the Superintendence of Electric Services, Gas and Telecommunications, to take on the supervisory roles of the government in the electric and fuel industry.

Decree-Law No. 11 of the Ministry of Economy

1984

Define fines and sanctions for electric companies.

Decree-Supreme No. 6 of the Ministry of Economy

1985

Define the norms to coordinate the operation of interconnected electric systems. Creates the CDEC and the rules for transmission pricing, rules modified later in 1990 and 2004.

Privatization

1986

Starts an extensive privatization process by which today almost 100% of the electric industry, generation, transmission, and distribution, are in private hands.

Decree-Law No. 119 of the Ministry of Economy

1989

Update fines and sanctions for electric companies.

Law No. 18,922 amends DFL No. 1 (Decree-Supreme No. 6) regarding electricity transmission pricing

1990

Improves the mechanism for electricity transmission pricing, introducing the concepts of area of influence, and a use and a capacity charge (tariff revenue and basic and additional tolls).

Decree-Law No. 327 of the Ministry of Mines that replaces Decree-Supreme No. 6 of the Ministry of Economy

1998

Improves the definition and exigencies of service quality and reliability, and the CDEC governance structure.

SIC and SING crisis

1999

Between 1998 and 1999 the SIC suffered a drought with a severe impact on hydraulic generation which has implied electricity rationing at the household level. In 1999 the SING suffered from severe blackouts because of the deficiencies in the coordination of the system.

Law No. 19,613

1999

Due to the severe drought that affected the SIC, the authority increases the control that SEC commands on the electric companies. Also, the droughts that happen to be more severe than the ones considered in the calculation of regulated prices are excluded to be considered as a cause of Force Majeure. (Continued )

Chile: Where It All Started

87

Table 3.4. (Continued) Event

Year

Comments

Resolution from the Ministry of Economy R.M. No. 88

2001

Generators are forced to supply all the electricity required by distribution companies at a price equal to the node price. The last independently of the generators’ electricity supply contracts with the distribution companies.

Change to DFL No. 1 by Law No. 19,940

2004

Changes in regulation to incentive additional investments in electricity transmission. Also changes the criteria to define regulated nodal prices; is enlarged the market for free contracts; is defined a mechanism to set network access charges; and are explicitly recognized the ancillary services within the law.

Natural gas crisis

2004 onward

Economic crisis in Argentina implied natural gas export constraints to Chile, which left the Chilean ESI operating at the limit of its capabilities.

Changes to DFL No. 1 by Law No. 20,018

2005

As a reaction to Argentinean natural gas exports constraints to Chile, regulation is changed to incentive additional investments, and to increase the reliability and flexibility of the electric system.

Main concerns with respect to the CDEC refer to: ●

●

●

●

●

the dichotomy that exists between the CDEC’s planning and dispatch of the generation– transmission system, that is made independently of the ownership and the generators’ electricity supply contracts; the collective responsibility that the law consign to the generators and the transmission companies while being required to operate their facilities according to the CDEC instructions; the large number of conflicts that existed between generators and transmission companies with respect to pricing, and energy and capacity transfers; the deficient information that exists with respect to the CDEC’s criteria to dispatch the electric system; and the influence that some agents can exert on CDEC’s decisions.

So far some progress has been made to solve some of these issues. First, in 2004, is the novel conflict resolution mechanism specially created for the electric industry. It consists of an independent board with seven members, who sanction on many of the conflicts that emerge within the electric companies and/or the authority.10 Second, and also in 2004, is the change in the criteria to define electricity transmission prices, from agreed tariffs to regulated tariffs. And third, in 2005, is the enhanced independence of the system dispatch and its operation, achieved by increasing the CDEC Board voting quorum required to remove the Executive Director. 10

In Law No. 19,940, the board is known as the Electric Experts Panel, “Panel de Expertos Eléctricos”, which is elected by the Chilean Free Competition Court.

88

Electricity Market Reform

3.2.2.2. Pricing For end users, the Chilean electricity price system is based on a tariff system designed by the CNE in 1980, before the reform, and formalized in 1982 by the new law, DFL No. 1. Since 1980 the authority recognized a market for small and one for large consumers, where in the beginning small consumers were identified as the ones with a load of at the most 4 MW. Later, the 4-MW threshold level is reduced twice, first to 2 MW in 1982 with the approval of the DFL No. 1, and then to 0.5 MW in 2004 with Law No. 19,940.11 Small end users buy electricity at regulated energy and power prices, while large end users are allowed to freely negotiate their electricity supply contracts. Final prices for small end users result from the addition of different charges: power (capacity or peak load), energy, transmission, and distribution; and more recently, ancillary services.12 Some of the costs are borne by generators and the others by distributors and end users through specially designed tariff formulas. Generators can sell electricity in three different markets: ●

●

●

A generators’ market which takes place in the CDEC, where generators trade to complete their electricity supply contracts. A market for large users who freely negotiate energy and peak load or capacity prices with generators and distribution companies, generally through long-term contracts. A third market is for small users, where distribution companies buy generators’ electricity and sell it to small consumers at regulated energy and power nodal prices determined by the CNE.

Transfer prices between generators. Generators trade energy and capacity in the CDEC. Energy transfers that result from the coordination of the system made by the CDEC are valued at the marginal costs of the system, calculated by the CDEC on the information declared by the generators. The CDEC’ marginal cost is built upon the marginal cost of the less efficient generator being dispatched. For transfers of peak power between the companies, the transfers are made at the power cost calculated by CNE every 6 months in the “Indicative Works Plan for Generation and Transmission”, by which it determines the least expensive generation–transmission expansion plan for the system.13 It happens that the CDEC determines the operation of the system independently of the ownership and the generators’ electricity supply contracts, which implied that generators’ energy and power transfers are cleared after they have signed their electricity supply contracts with large end users, distribution companies, or other generators. Thus, a generator

11

With these consecutive reductions in the threshold level, the number of end users eligible to bid their electricity supply contracts has been increased. 12 Law No. 19,940, March 2004. 13 Given a demand forecast for the following 10 years and accounting for existent and in-construction generation–transmission facilities, the CNE’s “Indicative Works Plan for Generation and Transmission” determines every 6 months the least expensive generation–transmission expansion plan. Before 1997, each generator was required to satisfy in advance an energy balance between its contracted energy and its firm energy, firm energy which is determined based on the generator’s installed capacity and expected available energy. In the energy balance the sum of the contribution of energy from a power plant and energy purchase contracts with other generating companies constitute the company’s firm energy. In 1997 the concept of firm energy that previously existed in the 1985 Decree-Law No. 6 disappeared, where in advance only a global energy balance should be satisfied by the system. Power transfers are determined based on a company’s peak power demand and its firm power supply, where firm power surplus are mandatory traded. Decree-Law No. 327.

Chile: Where It All Started

89

faces a risky future with no guarantee of being dispatch, as well as on the system’s marginal cost or capacity price at which the CDEC will clear the market. Here the computational model that determines the marginal cost of the system and the power plants’ dispatch comes as a fundamental piece of the puzzle;14 and during electric crisis or in circumstances when the system’s marginal cost abruptly increases controversies on the model and its parameter assumptions have provoked.15 Competitive prices for large consumers. By law, large consumers with a load above 0.5 MW contract electricity at prices, quantities, and conditions freely negotiated with the generators and distributing companies. The prices determined in these free contracts later play a key role in the small end users’ price; it is because by law they define an anchor for regulated energy prices for small end users, where small end users’ prices cannot differ more than 10% from the free prices.16 The main discrepancies with respect to free electricity supply contracts and pricing between large users and generators refer to service reliability and the responsibility of the different agents to ensure a secured electricity supply. Situations have been critical during energy crisis when the large end users have demanded from the generators and the transmission

14

In the 1980s and 1990s, the CDEC-SIC dispatch of the power plants used a model known as GOL (Gestión Optima del Laja), based on a marginal cost pricing criteria for electricity generation. The GOL model was inspired in Marcel Boiteux works what can be named as the “Electricité de France School” (see Boiteux, 1949, 1956, 1960); and also the GOL model follows the Jenkins and Joy (1974) model. Given water inventories, the expected hydrologies, hydro and thermal generation units, fuel and electricity rationing or failure costs, and projected demand, the GOL model minimizes the expected total costs, operation and rationing costs, to serve energy demand for a time period of 10 years. Until September 1991 the GOL model with quarterly periods was used to price energy transfers between generators. In October 1991 the model was replaced by the OMSIC model (OMSIC from its Spanish acronym for Operación Mensual del SIC, SIC’s monthly operation). This is a model with monthly periods that follows similar principles than the GOL model but that is specifically adapted to model the short-term operation of the system. More recently, the CDEC has replaced the OMSIC model used in the medium-term planning by the more sophisticated multi-dam multi-node model known as PLP. The CDEC-SING uses a comparable model, but without the complexity of having to model a hydro thermal system with an inter-annual dam. Rationing cost is understood to be the cost incurred in kWh, in average, by the end users when suffering from an energy shortage, and the energy shortage should be supplied with emergency generators, if it is so agreed. This rationing cost is calculated as a unique value, representative of the more frequent deficits than can show up in the electric system. 15 For example, controversies with respect to the criteria that are used to distribute capacity payments between generators: What is the contribution of each generator to supply energy during peak hours? What is the number of peak hours during which a power plant is eligible to receive a capacity payment? Should hydraulic generators receive capacity payments if they run out of water and cannot contribute with capacity? Should combined cycle natural gas power plants receive capacity payments if they run out of natural gas and cannot contribute with capacity? Is the capacity payment the right signal to assure enough peak load capacity for the system or it is given by the failure cost that the generators must paid to end users when they cannot satisfy the electricity demand? Of these questions, the number of peak hours that should be accounted for capacity payments was the first controversy solved by the Electric Expert Panel, “Panel de Expertos Eléctricos”, which was created in 2004. It define that for wintertime the number of peak hours that should be accounted for capacity payments is eight. 16 In 1982, DFL No. 1 circumscribes small end users’ prices to deviate no more than 10% from the free price. Afterward this band was changed twice. First in March 2004, Law No. 19,940 decreased the band from 10% to 5% and secondly in 2005, Law No. 20,018, the band changes in the opposite direction, by which the regulated energy price can differ from the free price up to 30%.

90

Electricity Market Reform

companies for additional investments and backup capacity; while the generators and transmission companies have demanded from the large end users, the installation of additional low-frequency (LF) relays to disconnect large blocks of consumption if a critical condition of operation in the electric system emerges. Regulated nodal prices. The DFL No. 1 established that, in electric systems whose installed generation capacity is above 1500 kW, the generation and transmission prices would be regulated when they supply electricity to small consumers or to distribution companies in the proportion that corresponds to small end users. The regulated nodal prices at the generation and transmission level are determined every 6 months by the CNE in April and October in the “Indicative Works Plan for Generation and Transmission”. This calculation is made based on a demand forecast of peak power and energy for the following 10 years, taking into consideration existent facilities and the ones in construction. The computational model is stochastic over a minimum of 40 hydrologies that until 1998 correspond to the 1941–1980 period. Under these set of assumption happened the worse hydrology corresponded to the 1968–1969 hydrologic year which later was exceeded by the drought that affected the country in 1998–1999.17 With these inputs, CNE determines the outcome that minimizes the present value of the expected cost of total supply, including expected rationing costs during the period of study. The basic energy price is determined as an average of the marginal cost of energy over the next 48 months of operation. The basic capacity price arises from the incremental cost of the more efficient generating units to provide additional power during the peak hours of the annual demand, increased in a percentage equal to the system’s theoretical reserve power margin calculated by CNE. The theoretical reserve power margin changes from time to time, and in April 2005, it was 11.76% for the SIC and the SING. Next and to account for energy and power prices for each node of the transmission network, the authority uses penalty factors which account for the energy and power losses of electricity transmission within the system. Finally, the regulated energy node price is checked against the competitive energy prices determined among generators and large end users, where previously calculated energy prices cannot differ more than 10% from the prices freely determined in the contracts between the companies and large clients. If the difference between the regulated prices and the free prices is greater than 10%, the CNE must multiply all the regulated node prices by a unique coefficient, higher or lower, to take the regulated prices within the band of 10% around the free prices. The CNE pricing model has been criticized because the parameters and assumptions do not reflect the conditions that frequently affect the industry. For example, in the late 1990s lack of investments in the SIC is attributed to the systematic reduction of nodal prices and the price rigidity has shown its inadequacy to signal the electricity crises. Both may help to explain the generators’ lack of interest to sign electricity supply contracts with distribution companies. Electricity prices. The path of regulated energy and capacity prices, and the CDEC spot marginal cost for the SIC and the SING are shown in Figures 3.5 and 3.6. 17

See Raineri and Ríos (1998); and Raineri (2000). The hydrologic data required was modified in 1999 with by Law No. 327 which requires a minimum of 40 hydrologic years, where on the April 2005 CNE Nodal Prices calculation, it used a total of 43 hydrologic years. The three additional hydrologies included in the model are: the first which roughly has 80% of the rain recorded in the 1968–1969 hydrologic year; the second with 90% of the rain recorded in the 1998–1999 hydrologic year; and the third that is more humid than the average sample, being determined in a way to keep the average of the hydrologic years sample rather constant.

91

Chile: Where It All Started SIC–SING energy price and CDEC marginal cost (million $ kWh, monthly average in April and October ) 160 140 120 100 80 60 40 20 0

October 2004

October 2003

October 2002

October 2001

October 2000

October 1999

October 1998

October 1997

October 1996

October 1995

October 1994

October 1993

October 1992

October 1991

October 1990

October 1989

October 1988

October 1987

October 1986

October 1985

October 1984

October 1983

October 1982

Energy SIC Energy SING MgC SIC MgC SING

Fig. 3.5. SIC and SING energy price and CDEC marginal cost. Source: CDEC, SIC, CDEC SING and CNE.

SIC–SING power price ($ kW month, monthly average in April and October) 18 Power SIC Power SING

16 14 12 10 8 6

October 2004

October 2003

October 2002

October 2001

October 2000

October 1999

October 1998

October 1997

October 1996

October 1995

October 1994

October 1993

October 1992

October 1991

October 1990

October 1989

October 1988

October 1987

October 1986

October 1985

October 1984

October 1983

October 1982

4

Fig. 3.6. SIC and SING power price. Source: CDEC SIC, CDEC SING and CNE.

In Figure 3.5 we see that since the 1980s exists a decreasing trend in the regulated energy price, decreasing trend which is steeper in the SING than in the SIC, where the regulated energy prices have come closer to the CDEC’s marginal cost. In the case of the SIC, the marginal cost depends on the hydrologic conditions, whereas in the case of dry conditions or a severe drought the system is forced to dispatch the more expensive thermal units. This explains the sudden increases in the system’s marginal cost for the 1996–1997 and 1998–1999 periods. In the case of the SING, the system’s marginal cost depends on the variable cost of its thermal power plants, which until the late 1990s depended on oil and coal prices, and since 1999 it also depends on the less expensive imported natural gas.

92

Electricity Market Reform

The idea of having stable end users’ prices was originally conceived in the DFL No. 1, for that we have, from Figure 3.5, that regulated energy nodal prices have followed a smooth and decreasing trend coming closer to the path that may well be predicted by the system’s dispatch marginal cost, which can be considered as the long-term average energy price. But besides the original intention of the authority, it happens that afterward and from the producers’ side, the regulated nodal energy prices resulted insensible to the short-term supply conditions that affect the systems; being a useless instrument to adjust supply and demand, a critical condition of operation happens in the system. The SIC and the SING regulated capacity prices are depicted in Figure 3.6. Conceptually capacity prices should be determined by the marginal cost of increasing the installed capacity, the marginal turbine which is determined by standard or novel technologies as was the case of the new natural gas power plants in the late 1990s. In both systems, the SIC and the SING, after the entry of natural gas, the capacity price decreased. But exist other events which can affect the regulated capacity price: the interconnection of isolated electric systems as it happens in the SING in the late 1980s; and changes in the theoretical reserve margin determined by the authority. The CDEC’s marginal costs of energy is shown in Figure 3.5, are monthly average, but the CDEC calculates hourly marginal costs, meaning that actual cost fluctuations are even larger, and are larger than the regulated energy price fluctuations determined by the biennial calculation of the CNE. After all and under the new regulatory framework which boosted competition between generating companies, in Chile we have decrease in electricity prices. To these also contributed the fine-tuning and learning process of the new regulatory model that followed its introduction since 1982. Chilean regulations have tried to promote private decisions based on economic signals through prices. Consequently, the developments of the electric and natural gas industry have been made by private agents, attracted in a competitive way toward the investment opportunities offered by the market. The new regulatory model and privatization of the most important companies at the end of the 1980s make possible large private investments in generation. It is in this new competitive atmosphere where the companies began to invest in thermal generation to meet the increasing demand of energy as well as the new environmental constraints. Based on these facts is that in the early 1990s the chances to import natural gas from Argentina emerged as the solution to the country’s energy needs. Electricity distribution pricing. The 36 distribution companies that exist in Chile have a nonexclusive public service concession over a geographic area and are mandated to serve the electricity demand at a regulated price called distribution value added or VAD. The VAD is a multiple-part tariff for using the distribution infrastructure and it is formed by: a fixed fee for managing, billing, and servicing the consumer; average energy and power losses; and a fixed fee per unit of power to pay for the operation, maintenance and investment costs. In VAD estimates, annual investment costs are calculated considering the replacement cost of efficient facilities for a projected demand, the facilities’ lifetime, and a 10% real return on assets. VAD is determined every 4 years for each distribution company based on a yardstick competition model where tariffs are defined according to an efficient model company that distributes electricity in the distribution company’s geographic area. Each distribution company’s VAD is determined by a cost study made by the CNE. The distribution company can contrast the CNE study with its own study, in which case VAD parameters result by weighting the CNE study on two-thirds and one-third for the company study. With the tariffs thus determined and the regulated nodal prices, the consistency of the distribution tariffs is

Chile: Where It All Started

93

revised for all distribution companies, where on the aggregate the average return on assets should be within a band of 10 ⫾4%. If the average return on assets is outside this band, tariffs should be adjusted such that the average return for all distribution facilities reaches the closer upper or lower limit of the band.18 Next, electricity price for small end users with a peak load of at most 0.5 MW is obtained by adding to the different VAD components, the regulated energy and power prices.19 The main problems experienced with electricity distribution pricing are as follows: ●

●

●

First, the weighting of the results, two-thirds for the parameter values obtained in the CNE VAD study and one-third for the parameter values obtained in the distribution company study, has a perverse incentive to create artificial costs structures which tend to diverge in the cost structure of the efficient model company. For example, past tariff processes have taken to differences in parameter values beyond 100% or even 200%.20 Second, the discount rate used to calculate tariffs is a fixed 10% real discount rate which may not reflect the opportunity cost of investments, and it should be replaced by one that reflects the opportunity cost of investments, as is currently done in other regulated industries in Chile where it is used as the capital asset pricing model (CAPM). Until 2004 electricity distribution pricing lacked a clear mechanism to set network access fees (access charges) for other agents who wants to sell electricity to large end users located in the distribution companies networks. Thus, only a few electricity supply contracts for large users located in the distribution companies’ networks have been served by a company different to the distribution company. This situation is expected to change with the new norms approved in 2004 that provide a clearer procedure to define network access fees.21

Electricity transmission pricing. The SIC developed to transport electricity from the hydro power plants located typically in the south of the country to the large consumption centers located in the center of the country;22 and the SING to connect the large mining 18

See Schleifer (1985) and Rudnick and Raineri (1997a). Before the regulatory change of 2004, Law No. 19,940, electricity transmission was paid by generators, where no explicit electricity transmission charge existed for small end users. Law No. 19,940 modifies that, where end users will pay 20% of electricity transmission and generators pay the remaining 80% through pricing formulas. Thus, in the near future, the structure of the prices for small end users will additionally consider a unique charge for transmission access in proportion of their consumption of energy, in such a way that the resultant price corresponds to the end user allocated cost at the level of production, transportation and distribution. 20 Some proposals have been made to replace the current mechanism (2/3 and 1/3 weighting factors) by one where, for example, the Electric Expert Panel, “Panel de Expertos Eléctricos”, chooses one of the tariff studies. See note. 21 Article No. 71-43 of Law No. 19,940 defines the criteria to calculate electricity distribution access charges for other agents who want to sell electricity in the distribution companies’ network. Basically, the fee is calculated in such a way that when the large user buys energy and power at regulated nodal prices, his final price, including the electricity distribution fee, should be equal to the price that a regulated small end user pays. The access charge is determined as an efficient component pricing rule for the distribution company. Today it is expected that with these access fees, competition to supply electricity to large end users within a distribution companies network increases as large end users start renewing their electricity supply contracts, and also because the number of end users who qualify to participate in the large customers market has increased. 22 In the same way as it happens with Brazil’s electric transmission system. In that respect, see Joao Lizardo R. and Hermes de Araújo (2006). 19

94

Electricity Market Reform

companies with the cities and power plants on the Northern coast. In both systems exists one private company (both Hydro-Québec subsidiaries) which owns the main transmission lines, frequently with a radial configuration. Until recently, revenues for transmission lines owners were determined by a multiple-part tariff approach defined by Law No. 18,922 of 1990, a change to DFL No. 1 in relation to transmission payments. According to it electricity transmission is paid by the generators as per their location within the network and as a function of their electricity injections and withdrawals from the network. Transmission charges consist of three components: ● ● ●

A “tariff revenue” (“Ingreso Tarifario”). A “basic toll” (“Peaje Básico”). An “additional toll” (“Peaje Adicional”).23

The “tariff revenue” is a variable fee associated to the differences in power and energy nodal prices within the different nodes and paid according to the generator injections and withdrawals of power and energy; while the “basic toll” and the “additional toll” are fixed fees calculated as a supplement to the “tariff revenue” to complete the payments that the transmission owner must receive, and they are paid depending on the generator’s area of influence and the location where the generator withdraws electricity.24 For a generator, the payment of the basic toll gives the right to withdraw electricity, without additional payments, in all the nodes of the system located within its area of influence. In addition, it gives the right to withdraw electricity, without additional payments, in all the nodes from which, in typical operation conditions of the system, net physical transmissions take place toward the area of influence.25 If the generator wishes to withdraw electricity in other nodes different to its area of influence, it must agree additional tolls with the transmission lines and substations’ owner.26

23

See Rudnick and Raineri (1997b). The total payment of electricity transmission facilities is defined as the annuity of the replacement value plus operation and maintenance costs. The area of influence should be understood as the set of lines, substations and other facilities of the electric system, direct and necessarily affected by the injection of power and energy by a generating unit. The basic toll is defined as a fixed fee that must be paid to complete the annuities corresponding to operation and maintenance costs and investment in lines, substations and other transmission facilities in the area of influence, after subtracting the tariff revenue. To this end, the basic toll is calculated in proportion to the maximum power transported by each user within the area of influence and must pay: the investments at their replacement value, assuming a lifetime of 30 years, the annual operation and maintenance costs, and a 10% real return on investment. The fees are revised every 5 years. 25 The net transmission, for these effects, is defined as the average energy transmission throughout a calendar year. This right will subsist in as much the net transmission stays toward the area of influence. 26 The additional tolls are calculated in the same form as the basic toll, and the payments give the generator the right to withdraw electricity in all the nodes located within the facilities. It also grants the generator the right to withdraw electricity, without additional payments, in all the nodes from which in typical conditions of operation of the system, net physical transmissions take place toward the nodes covered by the additional tolls. This last right exists in as much the net transmission condition indicated is satisfied. 24

Chile: Where It All Started

95

The main problems with electricity transmission have been the difficulties between transmission companies and generators to agree the basic and additional tolls, which can be explained because of: ● ● ●

the ambiguity of the concept of area of influence, a key ingredient to define the tolls; changes in the direction of net physical transmissions or energy flows; the vertical integration that existed in the generation–transmission segment.27

Together these have implied that some segments of the transmission system have been left unpaid, causing a negative effect on the incentives to invest in the additional transmission capacity required for the reliable operation of the system.28 In the year 2000, the vertical integration that existed in generation–transmission ended, but despite that, the disagreements about electricity transmission fees remained. To end with the controversies that existed on transmission fees the authority approved a change in the law in 2004, Law No. 19,940, which defines a complete new order for electricity transmission. Basically, it changed the model from one of the decentralize transmission industries with negotiated fees to one of the centrally planned segment or monopoly with regulated fees. The reform conveys: ● ●

●

27

the independence of transmission from generation and distribution; a change in the legal status of main transmission lines and sub-transmission lines to one of public service, which determines stricter service obligations;29 mandatory transmission investments defined in an electricity transmission expansion study supervised by CNE, which recognizes two types of expansions, new facilities and extensions of existent facilities;30

In the SIC until year 2000, ENERSIS S.A. controls: the largest generator ENDESA S.A., with 63% of generation capacity; the largest distribution company CHILECTRA S.A., with almost 100% of electricity distribution in the region where the capital city is located; and TRANSELEC S.A. owner of 98% of 500-kV transmission lines. The vertical integration ended in the year 2000 after ENDESA Spain obtained the control of ENERSIS holding, and it sold TRANSELEC S.A. to Hydro-Québec (see Raineri, 1999, 2003). 28 Since the early 1990s, many arbitrages have taken place to define basic and additional tolls; arbitrages which frequently resulted in contradictory results. The lack of technical and economic concepts of what is a generator’s area of influence explains most of the disputes. What was agreed in one arbitrage is not compulsory to what can be agreed in other arbitrages, implying that the concept of area of influence was defined in different and eventually contradictory manners in the different arbitrages. 29 The change in the legal status mean that the development and planning of electricity transmission turns out to be an activity in the benefit of end users and generators, and not as a generators’ development strategy or an extension of the generator’s activity to reach the consumption centers as was the case with the former law. 30 The study is made every 4 years. With that study as reference, every year the CDEC’s Operations Direction revises the consistency of the transmission facilities and makes a recommendation to CNE on the additional facilities needed. CNE, with the study and the CDEC recommendations, defines the expansion plan for the following 12 months. This expansion plan defines investments that are mandatory for transmission companies when they are an expansion of an existent facility, or investment projects to be publicly bid if they are a new development. The identification of the additional facilities, new or expansion arises from a regulated process with the representatives of all the players of the electric sector, and with regulated end users being represented by government authorities. For new investment projects, the execution and operation is carried out by the company that wins in an international public bidding. In the case of a development that is considered as an expansion of an existent major installation, the owner of the existent facility is the one that has the responsibility to carry out the works and operate the facilities in compliance with the law.

96

●

●

Electricity Market Reform the value to be paid for existent transmission facilities, which is determined every 4 years; and criterias to pay for the investments that finally are approved as necessary for the development of the electric system and to preserve its security.

The new tariff criterias for new investments discriminate between new facilities and extensions of existing facilities. For new facilities the tariffs consider the value of investment plus the operation, management and maintenance costs, and are applied during five tariff periods (for a total of 20 years), after which they are updated by the value determined in the main transmission study supervised by CNE. For investments projects identified as expansions of existing facilities the tariffs are revised every 4 years, which implies a larger uncertainty to recover investments as well as the operation, management, and maintenance costs.31 By law, the study which defines the expansion of the transmission system should consider existing generators’ facilities as well as the generation facilities that the companies declare in construction. Thus it happens that when a company declares the construction of a new power plant, it conditions the future development of electricity transmission, but in the end the company is not forced to build the power plant. What this means for the development of the electric industry is an open question, but so far the number of generators who has announced the construction of new power plants has increased compared to what was seen in the past years.32

3.3. Electric Crises The Chilean ESI has suffered three major crises since 1998: ●

●

31

In 1998 and 1999, the SIC experienced a drought that severely affected the availability of hydro resources. In 1999, the explosive growth of the SING installed capacity by the entry of new natural gas power plants led to a sequence of blackouts and brownouts whose origin was the lack of regulations which does not allow the proper coordination of a fast growing system, among them, and most importantly, rights were not defined for the appropriate provision of ancillary services.

With the new law, the trunk system is divided between the common area of influence and the rest of the trunk system. The common area of influence is paid in 80% by power plants injections and in 20% by consumption withdrawals (contracts). The rest of the system is paid by injection or withdrawal, completely, depending if the power flow gets in or out of the common area of influence. From this, it happens that, on average, the trunk system is paid 70% by energy injections and 30% by energy withdrawals. Tolls are defined according to the investments’ replacement value average costs determined by the CDEC Tolls Direction, according to the law and the proportional use that each generator’s injections make of the system. For sub-transmission, the tariffs are determined by the CNE every 4 years bearing in mind efficient investments at their replacement value. 32 With Law No. 20,018, approved in early 2005, that imply an increase in node prices, the generators’ interest to invest in power plants increased, where more than 5,600 MW of additional capacity for the next 10 years have been announced, demanding US$ 4 billion on investments. These announcements finally were included in the October 2005 CNE Indicative Works Plan for Generation and Transmission, where 5,700 MW of additional investments in generation are considered. This happened after a long period of time where generators expressed no interests to invest in generation.

Chile: Where It All Started

●

97

In 2004, the political instability in Argentina has exposed the excessive dependence of Chile on Argentinean natural gas, where the Argentinean natural gas export constraints have left the Chilean electric industry operating at the limit of its capabilities.

The causes of these crises, their effects, and efforts to address the problems are briefly described below. 3.3.1. First crisis The SIC electric energy crisis was triggered because in 1998–1999 the Central zone of Chile suffered the most extreme drought of recent years, which was complemented by numerous delays in the entry of a new natural-gas-fired combined cycle power plant (named Nehuenco), which would add 5% of additional generation capacity. The drought and the delay of Nehuenco, due to major technical failures, left the SIC with generating reserves below the minimum operating conditions. Under these circumstances the authorities approve three Electricity Rationing Decrees allowing electricity distribution companies to effectively ration electricity consumption in the months of March, April, and May of 1999. Fortunately by mid-1999, the hydrologic conditions change favorably and with it, the need to extend the third rationing Decree disappeared. The electric energy deficit engender both end users complain questioning the regulatory model and the generators’ disagreement on the price that should be used to pay for energy transfers between the generators: the price should be the system’s marginal cost or the system’s failure cost?33 Also, additional concerns existed with the use of the water that was stored in dams before the crisis, and that because 1998 ended with a rain deficit which was ignored. Regulated nodal prices are calculated with a stochastic model which consider a set of parameter values and assumptions, among which stands the range of 40 hydrologic years.34 On range of the 40 hydrologic years hydraulic generators made the case that droughts more severe than the ones considered in regulated nodal prices are circumstances of Force Majeure, which exonerates them from paying compensations to small end users as well to value the energy transfers that take place in the CDEC at the failure cost. Their main argument was that the applicability of the mathematical models should be restricted by the range of the parameters being used and model assumptions. Thus, the prices determined by the mathematical models should be applicable only if the system operates within the range of the parameter values being used. Their argument was backup because in a normal operation condition, small customers pay a risk premium within the regulated node price by which they will be compensated for the non-served energy. It is understood that the risk premium should be applicable only in the cases previously foreseen within the range of parameter values considered for the nodal prices calculation.35 Thus if the model is required

33

Traditionally the failure cost is far above the marginal cost of the less efficient power plant being dispatched, and the larger the energy deficit, the larger the failure cost. See Raineri and Ríos (1998); and Raineri (2000). 34 See Footnote 17. 35 The risk premium paid, the failure cost, forces the generators to compensate small end users for the non-served energy whenever electricity rationing occurs, in circumstances foreseen in the model and that cannot be invoke as Force Majeure. Because of that, the companies have the incentive to invoke as Force Majeure or Act of God those circumstances that were not considered in the calculation of node prices, as droughts more severe than the ones considered in the calculation of node prices.

98

Electricity Market Reform

to give price signals for all foreseeable and unforeseeable events, the range of the parameters considered should be wide enough to include also the extreme and less probable events that may occur. In the debate, thermal generators with surplus energy favor the idea that energy transfers should be paid at the failure cost as well as that the energy deficit should be equitably distributed among all the small end users of the distribution companies. Their incentives were to forward electricity to fulfill other generators – distribution companies’ contracts, being paid for that a price equal to the failure cost, well above the price at which they had signed electricity supply contracts with distribution companies. In addition, as the failure cost is increasing in the size of the energy deficit, it happens that the larger the deficit, the higher the price at which generators with surplus energy could sell their electricity. Hydro generators have been forced to take actions to diminish the negative adverse effects of the drought. In the case of ENDESA holding a predominantly hydro generator, it: ●

● ●

invest more than US $200 million in turbines, with an additional contribution of 651 MW of power; contracted 99.8% of the cogeneration capacity in operation; extended the transmission capacity from north to south to evacuate the generation capacity surplus in the North zone of the SIC.36

Colbún, another hydraulic generator and owner of Nehuenco, made significant efforts with plant builders to repair and put in operation the new natural-gas-fired combined cycle power plant as soon as possible. The reaction of the authority to the crisis was to change the law. In particular from these changes resulted the highly controversial change introduced in Article 99bis of DFL No. 1, where droughts, more severe than the ones considered in the calculation of node prices have been ruled out as causes of Force Majeure.37 Other changes in the law are: ●

●

●

energy transfers that take place in the CDEC should be valued at the system’s marginal cost, which for rationing hours was defined equal to the failure cost; the socialization of energy losses requiring the energy deficit to be equally distributed within the small end users of all the distribution companies; and a substantial increase on the sanctions for generators that fail to satisfy electricity supply contracts for distribution companies, and in the case of a Rationing Decree, rationing has to be executed proportionally to the commitments of each company.

The changes significantly increase the generators’ risk. They change the generators’ responsibility in the contracts with distribution companies as well as in the price that they have to pay for energy transfers that take place in the CDEC, risk that was not equally recognized within the regulated price. The flaws of the regulatory change undertook in Article 99bis became evident in the year 2001 when, after many purchase announcements, few distribution companies have fail in their intention to contract electricity from a generator, as required by law. Up to May 2005, the distribution company Saesa had made eight unsuccess purchase announcements; and CHILECTRA, the largest Chilean distribution company, also failed three times, and only 36

See Juan Eduardo Vásquez (1999). Article 99bis of DFL No. 1 was changed through Article 2 No. 2 of the Law No. 19,613 from June 2, 1999, published in the Official Newspaper on June 8, 1999. 37

Chile: Where It All Started

99

succeeded in 2003 after an additional change in the law was implemented, which constrained the generators’ liability for the non-served energy only to the true deficit. In 2001, the reaction of the government to solve the generators’ lack of interest to sign long-term contracts with distribution companies came through the Ministry of Economy, which on April 30, 2001 enacted the R.M. Exempt No. 88 by which it requires from the generator members of the CDEC-SIC to jointly satisfy all the distribution companies’ demand for electricity, independently of the distribution companies have or not an electricity supply contract.38 The decreasing trend of regulated nodal prices and the negative effect of Article 99bis underestimated the generators’ risk to sign electricity supply contracts, and that become apparent in the CNE’s 2001 SIC “Indicative Works Plan for Generation and Transmission” which did not project the entry of new power plants in the short term.39 At the time, the date of entry of the next power plant was predicted for January 2003, a natural-gas-fired power plant that later was not built, and in July 2003, a 570-MW hydroelectric power plant that finally began operating in 2004 with 640 MW. In terms of transmission projects, the “Indicative Works Plan for Generation and Transmission” expected for January 2004 an interconnection line between the SIC and the SING, which would contribute 250 MW to the SIC, and in July 2006 an interconnection line with Argentina which would contribute 400 MW to the SIC. So far, none of the transmission lines have been built. The SIC delay of investments in generation capacity results from the small capacity growth compared to the electricity demand growth (see Fig. 3.2). The lack of investments and the critical electricity supply situation expected for the years 2002–2004 concerns the authorities,40 who, in the desperation of avoiding a potential crisis, threatened the industry to purchase 250–500 MW of additional capacity if the private investors do not compromise these additional investments. Afterward the authority desisted when it was criticized by taking a role that is beyond its normative and supervisory roles. The main energy policy lessons from this crisis are as follows: ●

●

●

●

38

The lack of flexibility of regulated energy prices isolates small end users from large short-term fluctuations that happen in the market. The parameter values of the mathematical model can be exceeded in extreme conditions of operation, and require defining norms to operate under these conditions. Conditions which are excluded from being a cause of Force Majeure should be defined in advance and recognized in the prices. The industry should design a larger number of instruments which allows for a smooth balancing of supply and demand.41

In R.M. exempt No. 88 determines that the energy withdrawn by distribution companies without a contract should be paid at the regulated node price, as well as that the energy consumption should be proportionally distributed among all the generators. 39 The CNE Indicative Works Plan for Generation and Transmission is made taking into account all the power plants and transmission lines that the different companies have declared under construction. Thus, if the Indicative Works Plan for Generation and Transmission does not expect in the short- to medium-term new facilities, it must be because private agents have no interest to invest in the industry. 40 Between years 1999 and 2003, the installation of new power plants was expected to be inferior in 50% to what the authority anticipates that the demand will grow. See Magazine Qué Pasa, June 16, 2001. 41 Demand side management tools like tariff incentives and price–quality menus have been almost absent in Chilean ESI. The importance of price – quality menus for electricity supply contracts is analyzed in Raineri and Rudnick (1997).

100

●

Electricity Market Reform

The need to have a well-defined ancillary services market and its pricing mechanism, which may also consider payments for investments in backup capacity.42

3.3.2. Second crisis During 1999 the operation of the system faced two large disturbances on July 25 and on September 23, taking the system to blackouts, additionally to partial outages. Unlike the problems of the SIC which were caused by nature, the SING’s problems resulted from a sharp growth in the installed capacity (Fig. 3.2) that lacked the adequate change in norms and rules to guarantee the appropriate coordination of the large generating units and the large consumption of mining companies. The SING is a system where the large mining companies explain nearly 90% of the total energy consumption, which also occurs in large blocks with respect to total demand (the largest single mining consumption is 300 MW, the system’s peak load is 1567 MW, and new natural gas-combined cycle power plants are as large as 400 MW). A system of these characteristics becomes very unstable if occurs a sudden exit/entry of a large block of consumption and/or of a large power plant. Thus and given the particular characteristics of the system, in the heart of the crisis was the weak provision of the ancillary services being required to guarantee a reliable electricity supply.43 Contrasting with the SIC, the imbalances that cause the SING problems are not explained by the lack of energy, but have their origin in: ● ●

●

● ●

the uncoordinated entry and exit of large generating units; constraints in power plants’ static and dynamic properties (speed at which they can take or drop loads); constraints on the power plants’ technical minimum of operation and the system’s stability if it is operated with small loads with respect to the installation capacity; failures in the transmission system; and uncoordinated entry and exit of the large blocks of consumption from mining companies.

The large mining companies criticize the deficiencies of electricity supply and the ambiguity of the contracts, and inquire the generators to bear the costs of a deficient supply by: ● ● ●

●

investing in backup capacity; increasing spinning reserves; setting load shedding at 48 Hz to avoid the collapse of the system (in a 50-Hz system); and setting a limit on the maximum power that can be dispatched by each generator.

Although large mining companies recognized that in the year 2000 the incidence of blackouts diminished, they argued that the system is one that allows to fail and not to pay.44 As a consequence of the crisis, and to reduce the risk of future blackouts with a costly effect on large mining companies, and with the cooperation of member companies as well 42

Payments to backup capacity today are made through reserve margin payments. The margin of 15% was effective before 1998 and lowered to 6% in 2001 for the SIC and 5% for the SING, and lately they were increased to 11.76% according to the April 2005 Nodal Prices Decree for the SIC and the SING. 43 See Raineri et al. (2005) and Ríos et al. (2005). 44 See Electricidad Interamericana, December 2000 – January 2001.

Chile: Where It All Started

101

as large end users, since November 30, 1999, the CDEC-SING implemented a Short-Term Security Plan, which has had later revisions. In it, the CDEC defined the actions undertaken to mitigate the effects of a sudden disconnection of a large generating unit or a large consumption. On January 2001 this plan considered: ● ● ●

the installation of LF relays for a maximum of 340 MW; a margin or spinning reserves of 15%;45 and a limit on the maximum capacity to be dispatched by each power plant that today is of 250 MW.46

The above limit on the maximum capacity to be dispatched happens despite some of the newest power plants of the SING have an installed capacity of 400 MW. Thus, the question that come up is how the contribution of the different agents to the security of system will be paid, recognizing the fact that the new power plants have an installed capacity up to 400 MW but are restricted only to dispatch 250 MW. The SING’s crisis was essential to recognize within the Chilean legislation, the role that ancillary services have in the security and stability of the electric system, services that at the time of the crisis were voluntarily provided by few companies of the SING, as it also happens in the SIC. The more typical ancillary services required in the SING are voltage control, frequency regulation, primary and secondary injection of active and reactivate power, spinning reserves, and the backing up of power units when they are fine-tuned and run setup tests. Regarding the character of these services, some of them benefit all the agents in the industry, as a public good, and others only benefit a few, as a private good.47 The main energy policy lessons from this crisis are as follows: ●

●

●

The need to have suitable measures to allow the right coordination of all the agents who participate in the industry. The problem of the SING was not of an energy deficit, but rather it was a problem of a deficit of instantaneous power to absorb the sudden entry/exit of large blocks of consumption and/or large generation units. The need to have norms to backup the process of testing and fine-tuning of the new power plants. Particularly, in the current crisis this happened with new power plants whose capacity is from 20% to 52% of the peak load of the system. The need to have a well-defined ancillary services market and a pricing mechanism which adequately recognizes the character of each of the services. The effort must be placed on finding an incentives structure where the contribution to preserve the quality and the reliability of the electric service is a private and not a public good to which the different agents of the sector contribute in an unequal and non-remunerated basis.

Thanks to the SING’s Short-Term Security Plan, the failures of the system decreased from 192 in year 2000 to 72 in year 2003, where the ones that finally have had an effect on end users decreased from 65 in year 2000 to 13 in year 2003.

45

On October 2000, in the calculation of nodal prices, this theoretical power reserve margin was previously established in 5%. 46 First this limit was set at 180 MW, then it was increased to 200 and 220 MW in peak hours, and today it is in 250 MW. 47 See Raineri and Rudnick (1997); Raineri and Ríos (2001); and Galetovic (2001).

102

Electricity Market Reform

3.3.3. Third crisis From the early 1990s, the idea to import natural gas from Argentina matured. This country had important reserves and a growing infrastructure for its transportation, also complemented by the large natural gas reserves that were found at the end of the decade in Bolivia. Thus, the perception was one of the abundant natural gases in the region for many years. In the early 1990s the deregulation and privatization of the natural gas and the oil industry in Argentina facilitated in 1995 the natural gas integration with Chile, which allows the free import and export of this product between both countries. By the agreement, exports are allowed as long as there is no compromise to the internal supply of the exporting country, and in case of rationing it would be equally distributed within all the consumers independently of their country of origin.48 Argentina, as the exporting country, required that any company interested to export natural gas to Chile should prove that it had enough reserves to satisfy its domestic contracts as well as a surplus to satisfy its export contracts. Under this agreement, seven international gas pipelines have been constructed between Argentina and Chile. Under this situation, Chile began to import increasing quantities of natural gas from Argentina, since 1997 in the SIC and 1999 in the SING, whereas in 2003 Chile imported an average of 22 Mm3/day of natural gas from Argentina. In Chile 33% of natural gas imports are used for electricity generation, to fuel five thermal plants in the SING, with a capacity of 1443 MW49, and seven in the SIC, with a capacity of 2097 MW. The SIC natural gas power plants reach 23% of the installed capacity (see Table 3.3); and in the SING they reach 59% of the installed capacity (see Table 3.2). The SIC and the SING gas pipelines have fueled power plants, which between 1996 and 2004 represent 72% of the total increase in generation capacity (see Tables 3.2 and 3.3). The remaining natural gas imports are to supply the industrial consumption of a methanol plant (37%), and residential, industrial and State Oil Refineries consumption (30%). The introduction of natural gas has implied that the Chilean electricity sector became highly dependent on Argentinean natural gas, where in 2004, 28% of the SIC’s generation and 61% of the SING’s generation was based on natural gas. In this situation the system’s efficient dispatch significantly depends on the availability of natural gas, and any shortage of it can severely affect the cost of the systems and the availability of useful generation capacity required to satisfy electricity demand. Also, any shortage of natural gas puts in a risky condition; the payments being required to recover the large investments that have been made for its use. The success introduction of natural gas in Chile has been glimpsed in 2004 when Argentinean economic problems affect its capacity to satisfy the natural gas exports contracts. The economic imbalances developed in Argentina during the 1990s became unsustainable in 2002 with the devaluation of the local currency, which since 1992 was artificially set at one Argentinean peso per American dollar. With the devaluation, in a short period of time, the Argentinean peso dropped to a rate of 3 pesos per American dollar. The devaluation led to large changes on relative prices with significant adverse effects on income and employment. According to World Bank’s figures, between October 2000 and October 2002, the percentage of Argentinean population under poverty increased from 33% to 58%.50 To 48

Protocolo No. 2 de Integración Gasífera de 1985 (Natural Gas Integration Protocol No. 2 of 1985), as part of the Chile and Argentina Economic Complementation Agreement. 49 These power plants have been complemented with a 643-MW natural-gas-fired power plant (Salta), built in the northwest of Argentina, which through a 345-kV transmission line injects electricity into the SING. 50 World Bank (2003).

Chile: Where It All Started

103

diminish the short-term adverse effects of the devaluation on the population, government authorities pesified (fixed in pesos terms) the tariffs for basic utilities and transportation services, what differs to what have happened to other basic goods whose costs have risen by 30% on average. The pesified tariffs for natural gas consumption – household, transportation, electricity, and industry – become severely distorted with respect to the price of the closest substitute fuel. The artificially low natural gas price increased the domestic demand and also has replaced the use of the alternative but more expensive fuel. The pesified tariffs for the producers have a negative effect on the incentives to increase natural gas extraction, and to invest in exploration and the construction of additional transportation and distribution facilities. Thus sooner than later, production and transportation capacity constraints have become binding in a context of declining proven natural gas reserves subject to distorted energy prices. Argentinean natural gas exports mostly go to Chile, which represents 13% of Argentinean natural gas production. Thus, for Argentinean authorities, and if there is any shortage on natural gas supply, cutting exports to Chile appear as the less costly choice taking into account internal political constraints. Even though domestic tariffs in Argentina have been pesified, natural gas exports to Chile are under long-term contracts expressed in American dollars, meaning that natural gas arrives in Chile at prices roughly of US $2.5 MBtu (approximately US $1.6 MBtu at wellhead price), far above the Argentinean regulated natural gas prices for electricity generation of US $0.5 MBtu.51 As a result of the price distortions together with a modest recovery of the Argentinean economic activity after the 2002 crisis, Argentina has experienced a significant growth in domestic natural gas demand. However, due to the artificially low prices received by natural gas producers, their incentives to invest in exploration and the development of new gas fields and facilities has been eroded. From 2004 and with an increased emphasis in 2005, the Government of Néstor Kirchner in an effort to assure its domestic consumption has imposed severe constraints on natural gas exports to Chile, where exporting companies are mandated to fulfill domestic consumption prior to their export contracts. The natural gas deficit in Argentina that started in 2004 has constrained natural gas exports to Chile, meaning a deficit for Chile with respect to the consumption in a normal supply condition. Figure 3.7 shows that daily reductions on natural gas have reached levels of 50%, where the sectors most affected by these reductions are electricity generation and industrial processes. The thermal power plants that were able to substitute natural gas with an alternative fuel (diesel oil) have seen the fuels cost growing fourfold, affecting their relative efficiency in the CDEC dispatch and the system’s marginal cost. The industrial sector, when feasible, invested in backup capacity to continue with its operations. Household consumption was unaffected because Chilean authorities mandate that available natural gas should first favor households, hospitals, and small stores, and then electricity supply over large industrial customers. Looking back, the SIC and the SING benefit by having Argentinean natural gas at low prices. From its introduction and until March 2005 this implies savings close to US $3.5 billion on the fuel bill.52 However, since 2004, the natural gas deficit has left the electric system at the limits of its operational capabilities; the CDEC was forced to plan a more expensive

51

Crisis Energética, in Acontecer, Publication of the Instituto Tecnológico de Buenos Aires, May 1, 2004. Other benefits are the reduction in air pollution in Santiago, city that in the early 1990s was one of the most air-polluted cities in the world. 52

104

Electricity Market Reform

60

Natural gas injection constraints to Chile requested by Argentinean government respect to natural gas injections under unconstrained conditions. Total constraints between 05-05-2004 and 05-23-2005: 1328 Mm3.

50

%

40 30 20 10

2005-5-18

2005-4-6

2005-2-23

2005-1-12

2004-12-1

2004-10-20

2004-9-8

2004-7-28

2004-6-16

2004-5-5

0

Fig. 3.7. Imported natural gas constraints on Chile. Source: CNE.

dispatch using all the natural gas that is available at any moment, saving water in the case of the SIC for those periods when the natural gas deficit could be even larger. With this, the SIC’s reliability of electricity supply is left to what happened with the rain and hydro power generation. Natural gas supply in the SING depends not only on the decisions of the Argentinean authorities, but also on the Bolivian authorities’ decisions. This is because since August 2004 Argentina receives an important amount of gas from Bolivia in the north of the country (6.5 Mm3/day), through an agreement that expires in late 2005. The current situation implies a risky condition for the SING, because if Bolivia curtails natural gas exports to Argentina, the natural gas deficit in the northwest basin of Argentina would force its authorities to suspend all the gas delivered from that basin to Chile. In such a case, the SING remains with the generating capacity of other fuels, as coal, fuel oil and, diesel oil may be in fact not enough to satisfy the current SING’s electric demand. This situation illustrates that even though there are reserves and production capacity of gas in Argentina and Bolivia to satisfy the needs of the region, the geopolitical relationships are seriously affecting the Chilean electric market in the north of the country. The reactions in Chile to the Argentinean natural gas export constraints have been diverse, among which stand out the Chilean authorities lead to invest in a plant to imported liquefied natural gas (LNG). However, given the long lead times nothing is expected from this project until 2008, at the earliest. Also stands outs the changes in regulation that were approved on May 2005 which imply that:53 ● ●

53

natural gas shortage in thermal power plants cannot be invoked as a case of Force Majeure; generators are allowed to offer incentives to small end users to reduce electricity consumption when an energy deficit arises; Law No. 20,018, May 2005.

Chile: Where It All Started

●

●

●

105

distribution companies are allowed to sign long-term contracts, where the bid price will be considered in the nodal prices;54 a mechanism is defined to provide price incentives for investments in renewable energy; and until 2008, the generators who supply electricity to distribution companies (regulated end users) without contracts will receive a price which follows the system’s marginal cost.55

Even though some of these regulatory changes will not solve the short-term problem, they are expected to encourage new investments on a diversified energy matrix. At the same time regulatory changes have provided the generators with some instruments by which they can persuade small end users to reduce their consumption, with regulated nodal prices responding in a more flexible way to short-term supply conditions. The main energy policy lessons from the current crisis are: ● ● ● ●

●

an excessive dependence on natural gas imported from Argentina; the need to have end users’ prices responding to supply conditions; the need to have a diversified energy matrix; the need to have flexible tools or incentive instruments to balance electricity supply and demand in situations of extreme operating conditions; the need to have binding bilateral agreements.

3.4. Final Comments In the 1980s, Chile has pioneered the modernization and deregulation processes of the electric industry worldwide, where the deregulation of the Chilean electric industry responded to a change in the beliefs of the role of the State and the private sector on the economic activity. The Chilean electric industry followed the road of a market-oriented model. After more than two decades of experience with a deregulated and privately owned electric industry, the sector still goes under continuous adjustments which respond to a learning process of what have worked and what have not, as well as to the industry crises that have had significant impacts on the society. In this chapter we described the most distinctive features of Chilean ESI reform and three most important crises that it experienced since the late 1990s. The first crisis was in 1998–1999 when the SIC, a mostly hydraulic system, suffered a severe drought which took to an electricity rationing at the residential and industrial level; the second crisis was in 1999 that affected the SING, a thermal system, which experiences a fast growth on demand and generation capacity which was not complemented by the adequate norms to coordinate the operation of large new power plants and the large consumption of mining companies; and the third crisis which affected the Chilean ESI since 2004 occurred

54

The long-term contracts should not exceed 15 years, and the bid price has a 20% cap above the regulated nodal energy price. If in a contract a distribution company obtains a price which exceeds the average price of the industry by more than 5%, the difference that exceeds that 5% is socialized in the tariffs of all the distribution companies. 55 Basically, this refers to generators being mandated on April 30, 2001 by the R.M. exempt No. 88, by means of which it is ordered that generator members of the CDEC-SIC must supply all the electricity demand of distribution companies independently if they voluntarily have signed an electricity supply contract. The price would be equal to the nodal price plus some fraction of the difference between the nodal price and the CDEC dispatch marginal cost.

106

Electricity Market Reform

as Argentina has severely constrained natural gas exports to Chile, while the Chilean ESI has become increasingly dependent on natural gas imports from that country. The three crises imposed political pressures which ended up with changes in the law, changes that in some cases have been a response to the social concern that created the crises instead of a long-term vision of the rules needed to develop an efficient ESI. Looking back to what has happened and what has been made, it should be said that among the main energy policy lessons that can be derived from Chilean experience are: ●

● ●

● ●

●

●

● ● ●

the need to have flexible prices which do not isolate end users from short-term market conditions; the need of demand side management tools, like tariff incentives and price–quality menus; the mathematical models used to set regulated and simulated prices as well as for the system operation, reflect a circumscribed set of conditions, and this requires the definition of norms to operate under extreme conditions which exceed the ones considered by these models; the need to have a well-defined ancillary services market; the need for an adequate coordination among all the agents and a clear definition of rights and duties of each one; a clear definition of the reliability level that the ESI must have, and how much will be paid for that as well as for backup capacity; conditions which are considered as causes of Force Majeure should be defined in advance and recognized in the prices; the need to have diversified sources of energy supply; the need to have binding bilateral agreements; and the need to avoid regulatory uncertainty and to have stable rules with the flexibility to respond to the changes needed by the industry.

The Chilean ESI was among the first ones to be privatized and deregulated and after being battered for more than two decades, it offers many useful insights for policy makers and regulators around the world. Although Chilean electric industry have taken the road of market-oriented mechanisms, there is still some way to go to have a complete free market driven electric industry. Acknowledgments The author specially thanks the valuable comments made by Sebastian Berstein, Francisco Courbis, Juan Manuel Cruz, Raúl Espinosa, Luis Hormazabal, Eduardo Soto and the editors Fereidoon P. Sioshansi and Wolfgang Pfaffenberger.

References Decree-Law No. 327, 1998. Berstein, S. (1986). Tarificación eléctrica a costo marginal en Chile: Aspectos conceptuales, metodológicos y practices. IV Seminario Latinoamericano y del Caribe sobre Tarifas de Energía Eléctrica, Lima, Perú, November. Boiteux, M. (1949). La tarification des demandes en pointe: Application de la théorie de la vente au coût marginal. Revue générale de l’électricité, 58, 321–340.

Chile: Where It All Started

107

Boiteux, M. (1956). Sur La Gestion Des Monopoles Pubics Astreints AL’Equilibre Budgetaire. Econometrica, 24, 22–40. Boiteux, M. (1960). Peak load-pricing. Journal of Business, 33, 157–179. Comisión Nacional de Energía (1989). El Sector Energía en Chile, December. Decree-Law No. 6, 1985. Decree-Supreme No. 219, 1999. Decree-Supreme No. 287, 1999. Decree-Supreme No. 640, 1998. Electricidad Interamericana, December 2000 – January 2001. Galetovic, A. (2001). Asignación de costos debidos a centrales que operan para satisfacer restricciones técnicas en el SING, CEA, Departamento de Ingeniería Industrial, Universidad de Chile. Hachette, D. and Lüders, R. (1994). La Privatización en Chile. CINDE. Instituto Tecnológico de Buenos Aires, May 1, 2004. Crisis Energética, in Acontecer, Publication of the Instituto Tecnológico de Buenos Aires. Isaac, D., Santiago, A. and Eric, R.L. (2006). In this book the chapter Understanding the Argentinean and Colombian Electricity Markets. Jenkins, R.T. and Joy, D.S. (1974). Wien Automatic System Planning Package (WASP) – An Electric Utility Optimal Generation Expansion Planning Computer Code. OAK Rige National Laboratory, ORNL4945. Joao Lizardo R. and Hermes de Araújo (2006). In this book the chapter The Case of Brazil: Reform by Trial and Error. Joskow, P.L. (2001a). California’s Electricity Market Meltdown, April. Joskow, P.L. (2001b). California’s Electricity Crisis, March. Juan Eduardo Vásquez M. (1999). Energy Managing and Planning Director, ENDESA S.A., presentation at Pontificia Universidad Católica de Chile, September. Law No. 20,018, May 2005. Law No. 24,076, 1992. Regulates transport and distribution of natural gas in Argentina, authorizes natural gas exports subject to a government approval. Law No. 19,940, March 2004. Magazine Qué Pasa, June 16, 2001. Manifesto on the California Electricity Crisis, January 26, 2001. Institute of Management, Innovation, and Organization at the University of California, Berkeley. Ministry of Mines (1982). General Law of Electric Services, Decree-Law No. 1 of the Ministry of Mines (DFL No. 1). Natural Gas Integration Protocol No. 2 (Protocolo No. 2 de Integración Gasífera), 1985. Protocol That is an Element of the Chile and Argentina Economic Complementation Agreement. Public Emergency and Exchange Rate Regime Reform Law 25,561 (Ley de Emergencia Pública y Reforma al Régimen Cambiario 25,561), February 2002. Argentina. Raineri, R. (1999). Buscando el Control Corporativo: El Ingreso de Endesa España a la Propiedad de Enersis. Ediciones Universidad S.A. y M.N. Consulting Ltda., Santiago Chile, p. 138. Raineri, R. (2000). Comentarios al documento: Anatomía de una crisis eléctrica. Working Paper No. 101, Departamento de Ingeniería Industrial y de Sistemas, Pontificia Universidad Católica de Chile. Raineri, R. (2003). Becoming global: the entry of ENDESA Spain in the property of ENERSIS and the DUKE energy contest, Working Paper No. 142, Departamento de Ingeniería Industrial y de Sistemas, Pontificia Universidad Católica de Chile. Raineri, R. and Ríos, S. (1998). Costo de Falla y Precios para Valorizar las Transferencias de Energía en el CDEC, Working Paper No. 87, Departamento de Ingeniería Industrial y de Sistemas, Contract DICTUC S.A. – ENDESA S.A. Raineri, R. and Ríos, S. (2001). Contratos y Calidad de Suministro Eléctrico en el SING, Working Paper, Departamento de Ingeniería Industrial y de Sistemas. Raineri, R. and Rudnick H. (1997). Analysis of service quality standards for distribution firms. In L. Felipe Morandé L. and B. Ricardo Raineri (eds.), (De)Regulation and Competition: The Electric Industry in Chile. Ilades-Georgetown University, pp. 259–294.

108

Electricity Market Reform

Raineri, R., Ríos, S. and Vasquez, R. (2005). Business opportunities and dynamic competition through distributed generation in primary electricity distribution networks. Energy Policy, 33(17), 2191–2201. Ríos, S., Raineri, R. and Roca, M. (2005). Dynamic performance and resource mix modification in competitive environment. IEE Proceedings in Generation, Transmission and Distribution, 152(6), 770–779. Rudnick, H. and Raineri, R. (1997a). Chilean distribution tariffs: incentive regulation.In L. Felipe Morandé and B. Ricardo Raineri (eds.), (De) Regulation and Competition: The Electric Industry in Chile. Ilades-Georgetown University, pp. 223–257. Rudnick, H. and Raineri, R. (1997b). Transmission pricing practices in South America. Utilities Policy, 6(3), 211–218. Scheleifer, A. (1985), A theory of yardstick competition. Rand Journal of Economics, 16(3), 314–327. SEC Resolution 754, 2004. Subsecretaría de Combustibles de Argentina, 2004. Disposición 27/2004, GAS NATURAL. Approves the Rationalization Program for Natural Gas Exports and Use of Transport Capacity. World Bank (2003). Argentina crisis and poverty. World Bank Report No. 26127-AR, July 24.

Chapter 4 Electricity Liberalization in Britain and the Evolution of Market Design DAVID NEWBERY Faculty of Economics, University of Cambridge, Cambridge, UK

Britain was the exemplar of electricity market reform, demonstrating the importance of ownership unbundling and workable competition in generation and supply. Privatization created a de facto duopoly that supported increasing price–cost margins and induced excessive (English) entry. Concentration was finally ended by trading horizontal for vertical integration in subsequent mergers. Competition arrived just before the Pool was replaced by New Electricity Trading Arrangements (NETA) intended to address its claimed shortcomings. NETA cost over £700 million, and had ambiguous market impacts. Increased competition caused prices to fall, companies withdrew plant, causing fears about security of supply, but price–cost margins then increased and plant was returned to the system.

4.1. Introduction Britain’s electricity supply industry (ESI), among the first to introduce radical liberalization and re-organization, is one of the most studied models in the world. The initial England and Wales market design has been enlarged, has gone through several distinct phases since 1990. The Electricity Pool set up in 1990 was replaced by the New Electricity Trading Arrangements (NETA) in 2001, which was extended to cover Scotland under the British Electricity Trading & Transmission Arrangements from 2005, and continues to evolve as new rules are discussed and accepted. Figure 4.1 shows on a map of Britain the grid and main power stations connected at December 2004 (although the grid has hardly changed since restructuring in 1990, a large number of old coal- and oil-fired stations have closed and new gas-fired stations have opened). This chapter provides an overview of the developments of the Britain’s market, its evolution, and provides insights which may be useful to market designers in other parts of the world. The standard model of the ESI in almost every country before liberalization was an effectively vertically integrated franchise monopoly under either public ownership or cost-ofservice regulation. Investment in generation and transmission were (in theory) chosen to deliver the least-cost expansion plan (subject to government energy policy on fuel mix and 109

110

Electricity Market Reform

Fig. 4.1. Map of the UK with the location of the plants and transmission network. Source: NGC Seven Year Statements 2004/2005.

plant choice), financed by low-cost borrowing underwritten by the franchise revenue base. Britain was no exception, with the entire ESI under state ownership since nationalization in 1947. The Central Electricity Generation Board (CEGB) owned all generation and transmission in the whole of England and Wales, selling bulk power to 12 Area Boards, responsible for distribution and supply (retailing). In Scotland, the North of Scotland Hydro-Electric Board

Electricity Liberalization in Britain and the Evolution of Market Design

111

(NSHEB) and the South of Scotland Electricity Board (SSEB) each held regional franchises that included generation, transmission, distribution, and supply. The Government set the annual External Financial Limit restricting (publicly provided) borrowing, which in some years could be negative, implying a net dividend payment to the Treasury. The tariff structure was moderately sophisticated, with a two-part zonal Bulk Supply Tariff charging for capacity (of both generation and transmission), and variable costs (energy and regionally differentiated losses). Area Boards offered a variety of tariffs, with various forms of peak-hour capacity charges. While the pricing may have been sophisticated (although subject to government macro-economic considerations), investment planning, and particularly investment delivery, was poor, slow and costly, and there were few incentives to deliver cost efficiency. Liberalizing and restructuring the ESI was intended to replace this command and control structure with its cost-based (and often politically influenced) charges by a decentralized market-driven system that would nevertheless deliver secure, reliable electricity efficiently and at competitive prices. At the time the Government decided to restructure and privatize the ESI, there were few models available. The USA had evolved the contractual form of investorowned franchise monopolies under state cost-of-service regulation that was criticized for poor incentives, stranded investments and, in some states, high prices. Chile, on the other side of the planet, had been reforming and restructuring its ESI from 1978, and started gradually privatizing the sector from 1987. Norway already operated a spot market for wholesale energy for electricity generating and supply companies, but these remained publicly owned. With no obvious model to follow, considerable political pressure to deliver a competitive outcome (in contrast to the earlier privatizations of telecommunications and gas), and a tight timetable, the challenge was to design a set of markets and institutions to deliver these objectives. Of comparable importance, the design had to allow a smooth and predictable transition to a market-based system not just for electricity, but for the nationalized coal industry, three-quarters of whose (largely uneconomic) output was sold to the ESI, which in turn depended on coal for three-quarters of its output. The generation and distribution companies were to be sold to the general public and therefore needed predictable revenues on which they could be valued.

4.2. Restructuring and Privatization The debate on how best to restructure the ESI was vigorous, and hinged on the degree and nature of competition to be introduced. The CEGB unsurprisingly wished to remain monolithic, and argued that it would facilitate competition by new entry, but the Minister in charge of the process, Cecil Parkinson, was clear that he wanted to introduce more competition (Parkinson, 1992, Chapter 13). This required unbundling transmission from generation, splitting generation into several companies, and creating a wholesale market. In England and Wales this was achieved by separating out National Grid from the CEGB to create a transmission company covering England and Wales, but in Scotland the existing structure of two vertically integrated regional companies was retained, using the English wholesale market to introduce competition.1 Northern Ireland (restructured later) was different again in adopting the single buyer model as more suited to its isolation from the British Grid and very small size. 1

The Scottish argued that their system was smaller, with inadequate links to England, and worked well, so they argued for retaining the original structure.

112

Electricity Market Reform

The question of how best to introduce competition then resolved into how many generation companies to create, and this hinged on how to privatize nuclear power (and its investment program), an over-riding consideration of the Prime Minister, Margaret Thatcher.2 Privatizing nuclear power appeared to be inconsistent with more ambitious plans to divide the CEGB into as many as five to six generating companies. This led to a proposal to place the 12 nuclear stations in England and Wales with 8 GW into a large company, BigG, with the bulk of the fossil generation. The hope was that its resulting size would reduce the commercial risks attached to the nuclear assets and make the whole financially viable. It was then considered most sensible to group the remaining stations into a single company (LittleG) to countervail the might of BigG, rather than split them into several smaller companies. At a late stage the financial advisors made it clear the nuclear stations were not saleable at a positive price. The nuclear stations were transferred to Nuclear Electric and kept in public ownership. The Government thus divided the CEGB, with its 74 power stations and the National Grid, into four companies. Sixty percent of conventional generating capacity (30-GW capacity) was placed in National Power, and the remainder (20 GW) was placed in PowerGen. The high-tension grid, together with 2 GW of pumped-storage generation useful as rapid reserve,3 was transferred to the National Grid Company (NGC). The Electricity Act 1989 created the post of the Director General of Electricity Supply (the DGES), to regulate industry. This included enforcing and in due course modifying the price controls on the natural monopoly wires businesses of the NGC and the Regional Electricity Companies (RECs). He had a duty to ensure that reasonable demands for electricity were met, and that licence holders were able to finance their activities, to promote competition in generation and supply, to protect customer interests, and to promote efficiency. The Office of Electricity Regulation, Offer, was set up by the Government as an independent body under the Electricity Act, headed by the DGES. The four companies created out of the CEGB were vested (i.e. created) as public-limited companies (plcs) on March 31, 1990, at the same time as the 12 distribution companies, known as the RECs. NGC was transferred to the joint ownership of the RECs, and the RECs were sold to the public in December 1990. Sixty percent of National Power and PowerGen was subsequently sold to the public in March 1991, with the balance sold in March 1995. The pumpedstorage generation of NGC was separated and sold to Mission Energy at the end of 1995, and the RECs sold their shares in NGC in a flotation on the Stock Market, also at the end of 1995. The modern (English and Scottish) nuclear stations were finally floated as British Energy in 1996, with the rump of the original Magnox stations remaining in public ownership until decommissioned. Competition in generation was introduced, and all generators (public and private) were required to sell their electricity in a wholesale market, the Electricity Pool. The Scottish system, with about 10-GW capacity, was also restructured on March 31, 1990, when the NSHEB became Scottish Hydro-Electric, and the non-nuclear assets of the SSEB were transferred to Scottish Power. Both were privatized as vertically integrated utilities in June 1991, regulated on broadly the same basis as the industry in England and Wales. They were free to sell into the English market (and English generators were able to sell into Scotland and were entitled to access generation from the Scottish companies at the English Pool price, but in the early years prices were higher in England than in Scotland). Figure 4.2 shows the evolution of the UK electricity supply by fuel type (Britain accounted for 97.3% 2

Henney (1994) provides useful blow-by-blow account of events leading up to the restructuring decisions of 1988–1989. 3 Turbines pump water up to a hill-top reservoir during off-peak periods, allowing generation in peak periods or to provide rapid response to meet short falls in generation.

Electricity Liberalization in Britain and the Evolution of Market Design

113

350

300

250

TWh

200

150

100

50

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Net imports

Hydro⫹other

CCGT

Other steam

Coal

Nuclear

Fig. 4.2. UK generation by fuel type 1990–2004. Source: DTI (2005).

of UK capacity in 2003, with Northern Ireland making up the rest), while Figure 4.3 shows the capacity by fuel type of England and Wales (84% of total UK capacity). In 1990, the fuel mix was overwhelmingly coal (68%) and nuclear (19%), continuing the pattern of the previous decade, with no contribution from combined cycle gas turbines (CCGTs). Other noncoal steam (mostly oil) rose to 12% in 1991, but has rapidly diminished. Coal decline rapidly until 1999 when it fell below 30%. Gas-fired CCGT’s share rose rapidly, reaching 40% by 2004. Nuclear output peaked in volume (91 TWh) and share 28% in 1998, but has since fallen to 74 TWh and 21%. Clearly there have been dramatic changes in British electricity generation over the period since privatization. The prospectus on which the companies were sold (and their licences) also set out a timetable for introducing competition into supply. At privatization, the 5000 consumers with more than 1-MW demand were free to contract with any supplier (who could buy directly from the Electricity Pool), but all other consumers had to buy from their local REC, which had a franchise monopoly. In 1994 the franchise limit was lowered to 100 kW, and another 45,000 customers were free to choose their supplier. Starting in late 1998, the remaining 22 million customers had that right, and by mid-1999 the REC franchises finally ended. Table 4.1 below provides a timetable of major milestones in the British electricity industry (as does DTI, 2005). 4.2.1. Market and institutional design One of the most interesting institutional change in restructuring the British ESI was the creation of the Electricity Pool, a compulsory bulk electricity day-ahead market that determined the merit order and wholesale spot price of electricity in Britain. This operated as a compulsory

114

Electricity Market Reform 70,000

70,000

60,000

60,000

50,000

50,000 Peak demand 40,000

30,000

30,000

20,000

20,000

10,000

10,000

MW

40,000

0

Ap

ril Ap 199 ril 0 Ap 199 ril 1 Ap 199 ril 2 Ap 199 ril 3 Ap 199 ril 4 Ap 199 ril 5 Ap 199 ril 6 Ap 199 ril 7 Ap 199 ril 8 Ap 199 ril 9 Ap 200 ril 0 Ap 200 ril 1 Ap 200 ril 2 Ap 200 ril 3 20 04

0

Oil Coal Other OCGT CCGT others CCGT National power ⫹ PowerGen Nuclear

Import

Fig. 4.3. Plant capacity connected in England and Wales, 1990–2005. Source: NGC Seven Year Statements, various years, and data from Bower and Humphries.

day-ahead last-price auction with non-firm bidding, capacity payments for plant declared available (proportional to the loss of load probability (LOLP), and thus a negative exponential function of the reserve margin), and firm access rights to transmission (with generators compensated if transmission constraints prevented their bids being accepted). Each day generators bid their plant into the pool before 10 a.m. of the day ahead, and received their dispatch orders and a set of half-hourly prices by 5 p.m. for the following day. Bids had to be valid for the 48 half-hourly periods, although generators could specify various technical parameters (minimum load, ramp rates, etc.) in some detail to force a particular pattern of use over the day, and also influence whether the plant would set the price. The half-hourly system marginal price (SMP) was the cost of generation from the most expensive generation set accepted (including start-up costs where appropriate), based on a forecast of demand and ignoring transmission constraints. Generators declared available received capacity payments and, if dispatched, the SMP, which together made up the pool purchase price (PPP). All companies buying electricity from the pool paid the pool selling price (PSP) whose difference from the PPP was the uplift, which covered a variety of other payments made to generators. The Transmission System Operator (TSO, National Grid) used the same (rather ancient) software GOAL to dispatch plant as the former CEGB. As the successor companies had copies of GOAL, they could shape the rather complex individual plant bids (start-up, no-load, and three incremental prices plus various technical parameters) to optimize their revenue, rather than bidding the true parameters. The various institutions required to manage the decentralized system were codified in the Pooling and Settlement Agreement (PSA), a multilateral contractual arrangement signed by generators and suppliers which provided the wholesale market mechanism for trading

Electricity Liberalization in Britain and the Evolution of Market Design

115

Table 4.1. Major events in the British ESI. Event

Date

Comments

Electricity Act

1989

Provides legal framework for restructuring and regulation. Stephen Littlechild appointed first regulator (DGES) and sets up Offer.

Sale of RECs

1990

British ESI restructured, CEGB split up, Electricity Pool created, NGC transferred to RECs, RECs sold to public. 5000 sites above 1-MW customers free to buy in Pool.

Sale of generation

1991

60% NP and PG sold to public.

Coal Review White Paper

1993

Initial Coal contracts extended/replaced, 1990–1993 “dash for gas”, pit closures.

Sale of British Coal

1994

Coal industry sold to single private buyer.

Second-tier market

1994

100-kW market contestable (additional 45,000 eligible). Generators agree to divest plant within 2 years and accept a pool price cap.

End of Golden share

1995

Remaining 40% of NP and PG sold. Government Golden share in RECs expired, RECs subject to acquisitions.

First Price Control

1995

RECs subject to new price control, reopened after take-over wave.

NGC floated

1995

RECs float NGC, NGC’s pumped storage sold to Edison Mission.

British Energy

1996

Privatized.

Divestiture, bids

1996

Eastern leases 6 GW from NP and PG, NP’s and PG’s initial attempts to buy RECs denied.

Windfall tax

1997

£5.2 billion “excess profits tax” of privatized utilities levied by incoming Labour Government, value added tax (VAT) on energy reduced from 8% to 5%.

Regulation

1997

NGC’s price control reset, Government announces review of utility regulation.

Coal contracts end

1998

Contracts put in place in 1993 backed by REC contracts end, more pit closures, Government moratorium on gasfired generation.

Third-tier market

1998–1999

All 22 million customers contestable starting May 1998.

EU Electricity Directive

1999

EU Electricity Directive effective February, DGES report on pool prices (one of several). Merger of Offer and Ofgas into Ofgem.

Utilities Act

2000

Gas moratorium lifted November, continued generation sales.

NETA

2001

NETA introduced March 27.

British Energy collapse

2002

Low prices cause British Energy to enter administration.

Plant closures

2003

Plant mothballed following low prices, NGC predicts shortage, prices rise, plant returned.

Energy Bill, White Paper

2003

British Electricity Trading & Transmission Arrangements BETTA proposed, White paper on low carbon future.

ETS

2005

EU Emissions Trading starts January 1

BETTA

2005

British Electricity Trading & Transmission Arrangements go live April 1.

NP: National Power; PG: PowerGen.

116

Electricity Market Reform

electricity. It defined the rules, and required almost all parties wishing to trade electricity in England and Wales to do so using the Pool’s mechanisms. It provided the supporting financial settlement processes to compute bills and ensure payment, but did not act as a market maker. NGC owns and controls high-voltage transmission, and as the TSO was responsible for scheduling and despatch. Combining transmission with system operation has the advantage that the TSO can be provided with incentives for efficient operation that an Independent System Operator (ISO) would find too risky. With the sale of pumped storage, NGC no longer had any generation that could present conflicts of interest. NGC also acted as the Ancillary Services Provider, the Settlement System Administrator, and the Pool Funds Administrator, though again the provision of these services can be and often are separated from the provision of transmission services. In addition to the Pool, which acted both as a commodity spot market producing the reference price and a balancing market, most generators and suppliers signed bilateral financial contracts for varying periods to hedge the risk of pool price volatility. The standard contract was a Contract for Differences (CfDs) which specified a strike price (£/MWh) and volume (MWh), and was settled with reference to the pool price, so that generators were not required to produce electricity in order to meet their contractual obligations. These CfDs could be one or two sided, offering different hedging possibilities.4 Partly because the market structure was so concentrated, and partly because of the pass-through nature of the franchise contracts, other markets were slow to develop and remained very illiquid. The Electricity Forward Agreements market emerged as a screen traded over-the-counter market that allowed contracts to be traded anonymously and portfolio positions balanced. It failed to evolve into a futures market, partly because of the illiquidity caused by the large number of products (4-hourly periods for working and non-working days, for SMP, PPP, and uplift), but mainly because the underlying market was so uncompetitive. Contracts are not only important for risk sharing but were also critical in managing the transition from a franchise monopoly able to pass all its costs through to its captive customers to a market-based industry in which customers were free to buy from the cheapest supplier. The two major transitional problems facing the designers of restructuring were that British deep-mined coal was considerably more expensive than imported coal (and was soon to be revealed uncompetitive against gas), and that the CEGB had failed to set aside definable funds for decommissioning nuclear power plants. The surplus available to build up a decommissioning fund after paying for operating and fuel cycle costs were likely to be far too low given the likely equilibrium pool price. The first problem of transition was handled by a set of take-or-pay contracts between the generators and British Coal for the first 3 years at above world market prices. In parallel, the generators agreed contracts to supply the RECs for almost all their franchise output, for up to 3 years, that allowed the costs of the coal contracts to be recovered from these contract sales.5 There was the additional and very important benefit that the profit and loss accounts of the generators and RECs could be better projected, and these provided helpful financial assurance for the privatization to proceed. The second problem was dealt with by imposing a Non-fossil Fuel Obligation (NFFO) on the RECs (to buy electricity generated from non-fossil fuels, overwhelmingly nuclear 4

Over time the market developed quite sophisticated hedging instruments, for example hedging against prices above a specified strike price for the six most expensive half-hours in a month. The ability of the market to devise suitable hedges is relevant to the discussion whether regulators should insist on Reliability Options to protect consumers and reward rarely run plant (Vazquez et al., 2002). 5 The details of the various contracts required are set out in more detail in Henney (1994, pp. 120–124).

Electricity Liberalization in Britain and the Evolution of Market Design

117

power), and imposing a Fossil Fuel Levy (FFL) on all fossil generation (initially at the rate of 10.8% of the final sales price). This levy was paid to Nuclear Electric to build up a fund to meet its liabilities (of about £9.1 billion, which can be compared with the privatization proceeds from selling off the CEGB of just under £10 billion). There are two routes to effective competition in generation. The first and more satisfactory route is to ensure that capacity is divided between sufficiently many competing generators that no one generator has much influence over the price.6 This option was ruled out by the government’s initial decision to create one large company in order to privatize nuclear, and by the tight Parliamentary timetable which gave too little time to reconsider plans and to divide the generation companies further once it became clear that nuclear power was unsaleable. In the early years after privatization, the two fossil generators set the pool price over 90% of the time (the balance being mostly set by pumped storage, which arbitraged a limited amount of electricity from the off-peak to the peak hours). Nuclear Electric, Scotland, and France supplied base-load power that hardly ever set the pool price. Green and Newbery (1992) calculated that a duopoly unconstrained by entry would have significant market power and would be able to raise pool prices to very high levels (shown in Fig. 4.6). The second and indirect route to competitive pricing is to induce generators to sell a sufficiently large fraction of their output under contract, and expose them to a credible threat of entry if the contract price (and average pool price) rises above the competitive level. A generator that has sold power on contract only receives the pool price for the uncontracted balance. If this is a small fraction of the total (and it is usually about 10–20%), then there is little to gain from bidding high in the pool. High bids run the risk that the plant is not scheduled, leading to the loss of the difference between the SMP and the avoidable cost. The trade-off between lost profit on uncontracted marginal plant and higher inframarginal profits becomes increasingly unattractive as contract cover increases. Contracts and entry threats are complementary – entry threats encourage generators to sign contracts, and contracts facilitate entry (Newbery, 1998a). The advantage of creating sufficiently many companies for competition is that it does not need to rely on the continued contestability of entry, and it works well even when the competitive price is well below the entry price, in periods of excess capacity. As this route was not chosen, contracts and entry threats were all that remained, at least if price regulation was to be avoided. On vesting, the three generating companies were provided with CfDs for virtually their entire forecast franchise output, most for a period of 3 years. This both managed the transition to a free market and initially reduced their incentive to exercise spot market power to negligible levels,7 though not their ability to take advantage of transmission constraints and to game capacity availability. 4.2.2. Regulation of domestic suppliers, entry, and the “dash for gas” Initially only one-third (by volume) of the market was free to buy power in the Pool or by contract from competing suppliers, and the captive customers required regulatory assurance that their prices would be reasonable. This was assured by imposing price controls on the RECs. These allowed them to pass through the regulated charges for transmission and 6

One of the best ways to measure the extent of market power is the proportion of time that one generator is pivotal, i.e. whose supply is essential to meet demand, given all other sources of supply including imports. On this basis the two fossil generators were pivotal almost all the time for the first 5 years. 7 The obligation to take large amounts of coal under take-or-pay contracts also encouraged low bidding.

118

Electricity Market Reform

distribution, and the costs of purchasing power subject to a licence condition to purchase this “economically”. This raised the question of how own generation should be dealt with. Many of the RECs were keen to enter into generation on their own account, or to encourage it by others, not least to compete with the incumbent generators. If restructuring to deliver lower concentration had been ruled out at Vesting, then entry was even more critical to secure eventual competition. Indeed, one of the RECs, Norweb, had already begun to build a CCGT before Vesting, and the licences allowed the RECs to enter into generation up to a specified limit (15%) of total demand within their own area. Entry of new merchant independent power producers (IPPs) would be helped by the existence of long-term contracts for gas and electricity. Although there was a preference for truly independent merchant plant,8 there was now a precedent for RECs to enter into long-term power purchase agreements (PPAs) with IPPs, and to hold some equity in the IPPs. The PPAs allowed the IPPs to sign long-term contracts for gas (usually take-or-pay) and to issue comparable duration bonds. The economic purchasing requirement was intended to reduce the risk of “sweetheart deals”,9 and, critically, the franchise would end in 1998, limiting the possible damage to captive customers. Over the next 3 years substantial entry occurred. Before the end of 1990, contracts had been signed for 2.5 GW of plant (generally of 15-year duration) and by 1993 contracts had been signed for some 5 GW of gas-fired CCGT plant. This, in addition to the incumbents’ planned 5 GW of similar plant, would displace about 25 million tonnes of coal, or nearly half the 1992 generation coal burn of 60 million tonnes. The new CCGT capacity amounted to about one-sixth of existing capacity, which was in any case more than adequate to meet peak demand (although much of it was obsolete and was planned to be replaced by the CEGB with new nuclear stations). The “dash for gas” and the switch from coal more than halved the size of the remaining deep coal mining industry. The coal labor force had fallen from nearly 200,000 at the time of the 1984–1985 coal miners’ strike to about 70,000 by 1990, but further pit closures reduced numbers to 20,000 by 1993 and less than 10,000 by 1998. Figure 4.3 shows the rapid entry of gas-fired generation, and the resulting evolution of capacity connected to the National Grid. The decline in CCGT owned by PowerGen and National Power reflects industrial restructuring discussed below. Capacity payments were made to each generating set declared available for despatch, and were equal to the LOLP multiplied by the excess of the value of lost load (VOLL, initially set at £2500/MWh and indexed to the retail price index, RPI) over the station’s bid price (if not despatched) or the SMP (if despatched). This was set the day ahead and proved manipulable by declaring plant unavailable, and then re-declaring available on the day to collect the now raised payment. This practice was investigated by the regulator and new audit procedures were agreed to reduce the incentives for mis-reporting unavailability (Offer, 1992), together with new Pool rules for computing LOLP. This was now determined by the highest declared or re-declared capacity in the current and 7 previous days, so that there was an 8-day lag between declaring a plant unavailable and its impact on LOLP. A somewhat perverse implication was that the actual LOLP could be unity (certain power

8

Enron’s 1875 MW Teeside plant was proposed in 1989, and started construction in November 1990 (Henney, 1994, p. 222). 9 The DGES issued a statement to the Secretary of State on October 17, 1990 stating (inter alia) that “where a REC had taken equity in generation projects, I would need to be assured in addition that other generators, both existing and potential, had not been overlooked or put at a disadvantage.” (Reprinted in the privatization prospectus at p. 44.)

Electricity Liberalization in Britain and the Evolution of Market Design

119

cuts) while the value used to reward capacity could be almost zero. Newbery (1998c) argued that the computation of LOLP seemed excessive, given the high level of reliability over the first decade, and its overestimate may have contributed in part to the high-capacity payments. On the other hand, the VOLL seemed rather low, as Patrick and Wolak (1997) found that large consumers in one area were charged £7153/MWh in each of the three peak (or “triad”) half-hours in 1994/1995 for grid connection charges.10 As it was the product of VOLL and LOLP that determines capacity payments, these two possible errors may have been offsetting. In the 1994–1995 financial year, the generators earned £1421 million from capacity payments, or £24.5/kW/year, compared to £5821 million from selling at the SMP. Capacity payments were thus 20% of total payments for generation (excluding other ancillary services supplied by generators to the pool), and far higher than in earlier years when plant was more fully contracted. These capacity payments would have been sufficient to build 3 GW of new plant, or nearly 6% of total capacity. In the period 1995–1997, the annual average capacity payment was over £30/kW/year. During this period, the annual grid connection charge varied from £8/kW to ⫺£10/kW. The cost of keeping a new open-cycle gas turbine to provide reserve power might be £20/kW/year in interest and depreciation, and perhaps £6/kW/year for operating and maintenance (O&M) (MMC, 1996), so capacity payments should have been more than enough for security of supply. The exponential relationship between capacity payments and the tightness of demand (measured by the reserve margin) could also provide incentives for large generators to withhold plant, as Newbery (1995) demonstrated. Depending on the contract cover and plant margin, generators with a market share of about 30% might have an incentive to withdraw plant, exactly the opposite incentive to that intended. Green (2004a) examined the evidence and found that this strategy did not appear to have been significant. Later, dissatisfaction with capacity payments would be one of the factors causing the DGES to review the workings of the Pool and recommend the changes that resulted in the NETA of 2001, after which capacity payments were abolished. In addition to dispatching stations, NGC as TSO also had to resolve transmission constraints by paying out-of-merit generators to run if required (“constrained on”) or not to run (“constrained off”) in an export-constrained zone even if in the unconstrained dispatch. Under the vertically integrated CEGB the dispatch schedule automatically determined the security-constrained efficient dispatch, and the grid appeared adequately sized for such organizational form, but in the market-driven unbundled industry, the costs of resolving constraints rapidly increased (to £255 million in 1993–1994 or £4.3/kW/year). NGC offered an incentive deal to the RECs to share the benefits of reducing these and other costs, an idea that taken up by Offer, who was able to negotiate a better deal. The subsequent price control for NGC contained incentives to reduce constraint (and other ancillary service) costs, essentially by sharing the costs with a cap and collar. NGC proved adept at contracting for some plant behind constraints, making minor reinforcements to the grid, and scheduling maintenance to minimize these costs, reducing these constraint costs to less than 10% of their peak value.

10

Admittedly, these charges are not known accurately until after the peak, but large customers subscribe to moderately accurate forecasting services that can predict when prices are likely to be very high. The very low observed price response suggests that consumers value not adjusting the load in response to high prices, and by implication attach an even higher value to not losing the load.

120

Electricity Market Reform

Another criticism of the Pool was that it was only half a market, lacking any demand-side bidding. That was not quite correct, as NGC operated an annual tender auction for the provision of standing reserve to assist in its system management function. Standing reserve was provided by open-cycle gas turbine and pumped-storage plant, but also by demand reductions and non-centrally despatched small generators, though all had to offer amounts in excess of 3 MW. Large consumers could therefore specify their availability and willingness to reduce demand in various seasons and at various times of day, and NGC then accepted bids for which the total cost of providing load reductions were less than VOLL. In 1997/1998, 1809 MW of centrally despatched generation and 458 MW of demand modification and small-scale generation were contracted (NGC, 1997). The offer curve of such bids suggests that while there was some moderately cheap demand-side flexibility, beyond a quite modest level consumers needed a higher value than VOLL to be willing to curtail load, again suggesting that VOLL may have been underestimated, and that short-run demand elasticities for electricity were very low (with current control and metering devices).11 In addition, the Pool developed a less successful form of demand-side bidding directly into the Pool, and again its failure was an additional source of pressure to reform the Pool. The PPP determined the price of raw (unconstrained) energy and capacity, but generators and consumers are interested in the price at their location. It was appreciated that the theoretical solution to efficient spatial pricing is locational marginal pricing (LMP) developed by Bohn et al. (1984). As NGC, encouraged by Offer, explored a more satisfactory solution to the rather hastily designed system put in place at privatization, it was recognized that LMP faced a number of potentially serious drawbacks, not least of which was that its performance in the presence of considerable market power was untested (and largely unknown). The additional basis risk of trading at a large number of grid points whose price could diverge considerably from the pool price would require a large number of potentially illiquid contracts to cover risk. The possible gain in allocating the costs of transmission constraints and losses more precisely was not thought worth the loss in transparency and market liquidity. NGC therefore retained zonal access charges based on the incremental costs of reinforcing the grid to meet demands and supplies in that zone. NGC also publishes annual Seven Year Statements (looking ahead 7 years) which update predictions of demand and supply by zone, and indicate where new generation might best locate. The transmission charges are paid by consumers based on demand at the three half-hours of system maximum demand separated by 10 days (the “triad”), and by generators based on transmission entry capacity (or output in the triad if facing a negative grid charge). The more serious weakness in locational pricing was that, in contrast to the CEGB period, transmission losses were not borne by generators, distorting the merit order, while firm access rights rewarded, rather than penalizing, generators in export-constrained zones. Scotland was the obvious example of both problems, and two successive attempts by Offer to introduce transmission losses were successfully appealed to the courts. Green (2004b) estimates that nodal rather than uniform pricing in the presence of market power would have raised welfare by possibly 1.8%, which is high compared to the gains of restructuring discussed below. 11

Enthusiasts continue to believe that low-cost ICT (Information and communication technology) will enable even domestic consumers to time-shift loads such as freezers, hot water and storage heaters, and air conditioners where their thermal inertial allows electricity to be stored for modest periods in the form of heat (or cold). Evidence that this is cheaper than carrying generation reserves remains sparse, although Italy has now installed over 23 million automated meters which have the capability to manage domestic load (see Automated Meter Management Presentation to CRE, Michele Mazola, Paris, June 16, 2005, IBM).

Electricity Liberalization in Britain and the Evolution of Market Design

121

4.2.3. Performance after privatization Privatization and restructuring the CEGB delivered substantial improvements in efficiency, as Newbery and Pollitt (1997) document. They estimated that after the first 5 years, costs were permanently 6% lower than under the counterfactual continued public ownership, with a present discounted value at the public sector discount rate of 6% equal to a 100% return on the sales value of £10 billion. Labor productivity doubled, real fuel costs per unit generated fell dramatically (even in the publicly owned nuclear company), and substantial new investment occurred at considerably lower unit cost than before privatization. The contrast with Scotland was striking, where a similar social cost–benefit study by Pollitt (1999) found negligible efficiency improvements. One reason was undoubtedly that the two Scottish companies were not restructured, and remained vertically integrated, making it more difficult for competitors to gain access to their home market, even though nominally Scotland was able to trade in the English Electricity Pool. Scotland was an exporter through a severely constrained interconnector that was not efficiently priced, and had only two local generators, reducing the prospects of competition. Figure 4.4 shows the average price of domestic electricity in Edinburgh, Scotland, and London, England. Initially, London was 10% more expensive than Edinburgh, but by 2001 Edinburgh was almost 10% more expensive than London. This raises the question on how access to these scarce interconnectors should be determined and priced (NGC, 2004). If full nodal pricing is thought problematic, then “market splitting”, in which the System Operator (SO) determines when constraints isolate markets, and then sets market clearing prices in each zone, as in Norway, would seem attractive. In particular, it would allow English generators to contract with Scottish consumers, and this counterflow would release more export capacity from Scotland, as the constraint only applies to net electricity flows. English generators would effectively be paid to export to Scotland an amount equal to the excess of the English marginal price over the Scottish marginal price

12 Edinburgh London

Pence/kWh (2003 prices)

11

10

9

8

7

04

03

20

02

20

01

20

00

20

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

19

19

90

6

Fig. 4.4. Domestic electricity prices at 2003 prices excluding value added tax (VAT). Source: DTI Energy Prices, various issues. Figures are averages for credit customers taking 3300 kwh/year.

122

Electricity Market Reform

(which should include transmission losses), and would thus be able to compete effectively in that market. Under the existing system Scottish generators could price locally up to the English Pool price, as the shadow price of the export constraint was not made explicit and they did not pay for the quite substantial transmission losses.12 In the event the British Electricity Trading and Transmission Arrangements (BETTA) took the simple (if not efficient) approach of just absorbing the interconnectors into the regulated transmission system, appointing NGC as the GB System Operator, and treating the whole of Britain as a single market (which, as noted above, in contrast to nodal pricing, is costly).13 The lesson that vertical unbundling (at least legal, and preferably of ownership) is essential for effective competition has been accepted in the new EU Electricity Directive, and in the consultation for BETTA that started in 2003. Privatization, combined with unbundling and a transparent wholesale market, provided incentives for considerable efficiency improvements, but the concentrated market structure initially allowed the incumbent generators to retain these cost reductions as enhanced profits. The social cost–benefit analysis of Newbery and Pollitt (1997) found that while the overall simple sum of net benefits of privatizing the CEGB was nearly £10 billion, consumers lost relative to the counterfactual in which fuel prices fell and the CEGB had set prices as in the past, while the owners of the generation companies gained very substantially. Unbundling the natural monopoly transmission and distribution functions also allowed incentive regulation. Transmission and distribution companies are now subject to price controls, reset every 5 years. The price cap is indexed to the RPI with a projected productivity gain, X, hence the short-hand description of RPI-X, meaning that an index of regulated prices can increase by no more than the percentage increase in RPI less X each year of the control period. The initial level of the price index, P0, is also periodically reset by the regulator, Offer (subsequently Ofgem). In contrast to tradition cost-of-service regulation, price caps offer the incentive to cut costs and keep (at least until the next price control) the resulting increased profits. Recent price controls have employed benchmarking to establish the efficient frontier, further improving incentives. Improvements in the first 5 years under the initial price controls were modest as the X factor set at privatization was unambitious. Most of the price cuts, efficiency gains, and transfers to consumers took place after the price controls were reset (and after much merger and acquisition activity). After an initial increase in controllable costs of some 15% in the first 4 years after privatization, these costs fell by 30% from 1994 to 1998, and labor productivity doubled. The efficiency gains through cost reductions and investment savings amounted to £7.1 billion at a real discount rate of 6%, assuming a counterfactual productivity growth of 2% p.a. This was the target set in the public sector although performance in the 4 years before privatization was closer to 1% p.a. The restructuring was costly (at £1.1 billion) so the net gain was £6.1 billion (rounding), comparable to the gains of restructuring the CEGB. Customers gained £1.1 billion, the Government lost about £4.5 billion in future dividends but sold the companies for £8.2 billion, while the buyers gained £9.5 billion in future dividends for their purchase (all discounting at 6%) (Domah and Pollitt, 2001). 12

Marginal transmission losses from Northern generators to the load centers were often greater than 10%. Despite various attempts and judicial review Ofgem “is of the opinion that it is not legally possible for it to approve this Modification Proposal” (to introduce cost-reflective charging for transmission losses). Ofgem’s Information Note of January 30, 2004. 13 Under the Energy Act 2004 the Secretary of State amended transmission license conditions in September 2004 to create a single GB-wide set of arrangements for trading energy and for access to and use of a single GB transmission system. This came into effect on April 1, 2005.

123

Electricity Liberalization in Britain and the Evolution of Market Design 7000

35 Restraint

30

Profit maximizing Tacit collusion

6000 Plant withdrawal

£(2001)/MWh

25 20

5000 4000

NETA

3000

15 Electricity

10

2000

Coal cost Gas cost

5

HHI

Price control

1000

Coal HHI 0 04

03

20

02

20

01

20

00

20

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

19

19

90

0

Fig. 4.5. Real wholesale electricity and fuel prices 1990–2005. Source: Pool data and UKPX RPD data, DTI Energy Prices for fuels, HHI from Bower and Humphries.

Figure 4.5 summarizes a long and turbulent period of pricing in the England and Wales wholesale market, during the entire life of the Pool until 2001, and under the NETA thereafter. Hourly and daily price volatility was very considerably higher than the smoothed figures shown. Averaged over the year 1997/1998, for example, the average spread between the highest and lowest half-hourly PPP prices on a day is 180% of the average price on that day, and the standard deviation of half-hourly prices over the year is 78% of the average PPP.14 Figure 4.5 shows the fuel cost of generating electricity from coal at 36% thermal efficiency and from gas at 50% gross efficiency (55% net efficiency), and hence the margin between the yearly moving average wholesale price and avoidable cost. The line with diamond markers gives on the right-hand scale the Herfindahl Hirschman index (HHI) of market concentration of coal-fired plant (for most of this period the price-setting plant). This is the sum of the squared percentage shares of available capacity, so that the initial value of just over 5000 represents the equivalent of a duopoly.15 The evolution of prices is divided into periods identified by Sweeting (2001). To understand them it is first necessary to discuss the determinants of imperfectly competitive equilibrium prices in an electricity pool. 4.3. Characterizing Market Equilibrium in a Pool Modeling price formation to understand market power and market efficiency is a challenging problem that is not yet fully solved. Green and Newbery (1992) modeled the English Electricity Pool by adapting Klemperer and Meyer’s (1989) supply function equilibrium 14

Some of this variability is predictable and can therefore be hedged. Characterizing the unpredictability of prices, which is a measure of risk, requires correcting for predictable time variations over the day, week and year. The standard deviation of the difference between the actual half-hourly price and the moving average for that hour was £14/MWh or 55% of the average price. 15 The number of equivalent firms is 10,000/HHI.

124

Electricity Market Reform Feasible supply functions Duopoly and Quintopoly 100 2-firm range 5-firm range

GBP/MWh

80

D

Demand variation

Maximum demand C

Cournot line

60

B

40 20

Marginal cost

X

A

0 0

5

10

15

20

25 GW

30

35

40

45

Fig. 4.6. Equilibrium price range ignoring entry threats and contracts. Source: Calibrated for England 1990.

(SFE) model. The model is difficult to solve and typically gives a continuum of equilibrium prices. Figure 4.6 reproduces their calibrated model for England and Wales, ignoring contracts and entry threats. This approach is attractive and appears to be supported by companies’ claims that they bid supply schedules. It assumes a single-price gross pool with bids that hold for a reasonable period of time over which demand varies, as with daily bidding in the English Pool. In its simplest form it assumes that the supply functions bid are continuous and differentiable, and that demand is linear with constant slope but varies over the 48 half-hours. Each generator chooses a supply function that maximizes his profits given the residual demand he faces, made up of the variable total demand less the total supplies bid by other generators. As his bid has to be valid over the whole daily range of residual demands, instead of choosing a single quantity to submit to the market that would determine a single price (as under the Cournot assumption), he has to choose a continuous function relating the quantity that he is willing to offer at each price realization. The set of feasible solutions will be Nash equilibria in supply functions. There are, however, difficulties with this approach. Most pools (and the English Pool in particular) restrict bids to a single price for each quantity offered, producing a step function or ladder rather than a continuously differentiable function. The Amsterdam Power Exchange is a good example, where their web site provides the bid and offer ladders for each hour, their intersection providing the market clearing price for that hour. Fabra et al. (2004) argue that this radically alters the nature of the equilibrium, and requires modeling the market as a last-price auction, following on the earlier article of von der Fehr and Harbord (1993). They solve this if there is a single period and a known inelastic demand (up to a binding price cap), but cannot characterize the solution for bids that must hold for many periods (48 in Britain) with uncertain or varying demand. Hortacsu and Puller (2004) use data that is in step function form, which they then smooth to determine the marginal revenue of the residual demand facing each generator, and demonstrate that at least for the larger companies their bids appear to be profit maximizing against this smoothed schedule. Newbery (1992)

Electricity Liberalization in Britain and the Evolution of Market Design

125

suggested that if generators randomized over the positions of the steps in a step function, they could replicate a differentiable supply function, but it remains an open question whether this would be an optimal response to such behavior on the part of other generators. Standard Cournot oligopoly models are simpler, can be defended in tight market conditions, but suggest a more deterministic outcome than supply function models with their range of indeterminacy. Increasingly, consulting companies are developing price-formation models, the best of which capture the strategic aspects of supply function models with more careful modeling of the non-convexities of start-up costs which can dramatically influence the cost of providing additional power for short periods. Despite this apparent diversity of approach and the rather unsatisfactory theoretical foundations of bidding models, the evidence from various markets is consistent with the SFE story. Competition is more intense (closer to Bertrand) and prices closer to avoidable costs with spare available capacity, but as the margin of available capacity decreases, competition becomes less intense and outcomes closer to Cournot (as in the SFE). However, there remain two additional considerations before we can understand the English price evolution shown in Figure 4.5. First, while the possession of market power is legal, abusing it is not, and dominant generators need to be aware of the threat of competition references. That is the simplest explanation of incumbent bidding behavior from 1990 to 1994, where an acceptable level of prices at which to aim was arguably the entry price. The second important feature of the Pool is that it is a repeated auction, repeated every day and with the evidence of bids and outcomes available with a relatively short lag to the participants. The European Commission provides a characterization of collective dominance as a situation in which the market characteristics are conducive to tacit co-ordination and such co-ordination is sustainable, that is it is profitable and deviations can be deterred. The market characteristics that are conducive to tacit co-ordination include concentration, transparency, maturity, with a homogenous product produced by companies with similar costs and market shares, facing an inelastic demand, and with barriers to entry. Evidence supporting such a finding would include excess price–cost margins, profits, and an insensitivity of prices to a fall in cost. With the important exception of barriers to entry, the English Pool appeared to have all these defining characteristics and behavior. Tacit co-ordination was therefore to be expected, and market surveillance should clearly take account of this possibility. 4.3.1. Tacit co-ordination in the Electricity Pool Sweeting (2001) tested for tacit co-ordination by looking at individual company bids, subtracting all other bids from total demand to determine residual demand, and asking whether the bids were profit maximizing given the residual demand (but ignoring contract positions). He finds that in the first period up until 1994 both incumbents bid less aggressively than would be (short-run) profit maximizing, and that the prices were on average around the level at which entry was just profitable. If we were to conjecture on what strategies collectively dominant incumbents might co-ordinate, given close regulatory scrutiny, then keeping the price at the entry level while dividing the market in proportion to some objective criterion (such as plant capacity) would be plausible. It would also explain why both companies were keen to build new CCGTs even when their economics were marginal,16 for 16

While it is true that IPPs also entered, they did so on rather favorable long-term contracts not available to the incumbent generators. Certainly given the early gas prices and CCGT efficiencies, and compared to the opportunity cost of coal, the economics of investment were very marginal, as the House of Commons (1993) argued.

126

Electricity Market Reform

this would allow them to justify an increased market share (or in practice, in this prisoners’ dilemma, maintain market share in response to investment by the other company). As Figure 4.5 shows, fuel costs continued to fall away from electricity prices, but the generators nevertheless made several major increases in electricity prices, to the point that the regulator claimed that they were excessive (compared both to avoidable costs and the cost of new entry). After discussions with the regulator (and threats of a reference to the Monopolies and Mergers Commission) the companies agreed to sell 4–6 GW of plant within 2 years, in order to facilitate competition, and agreed to price caps (on both the annual average demand- and time-weighted pool price) for that period. The companies sold 6000 MW to Eastern (later TXU) with an earn-out clause of £6/MWh,17 ostensibly to compensate for the sulfur permits transferred with the plant and to reduce the buyer’s risk, but with the additional consequence of raising their rival’s marginal cost when bidding into the pool. Sweeting found that during the period from 1996 (after divestiture when the price cap ended) to 1998, each company’s bids seemed to be best (i.e. individually profit maximizing) response to those of the other companies and thus each firm was non-collusively maximizing profits. The price–cost margin increased as the regulatory threat of market abuse was replaced by (rather relaxed) competitive pressure, and the incumbents were probably quite happy to have sold plant at prices reflecting market power, in a market that was continuing to experience rapid entry. The ability to sustain a high price–cost margin depends on the volume of excess capacity, which was threatening to increase rapidly unless more coal plant were withdrawn or scrapped. At the same time Offer and Parliament (through the selection committee that investigated the energy industries) were becoming increasingly convinced that the Pool was not working well, and that the detailed rules of the dispatch algorithm were being manipulated to increase profits (Offer, 1998a–c). Henney (2001) notes other sources of discontent, notably the Labour Party’s belief that the Pool used “an operating and pricing system that was not competitive and was weighted against coal” (Robinson, 2001).

4.4. The NETA In October 1997, the Minister for Science, Energy, and Technology asked the DGES to review the electricity trading arrangements and to report results by July 1998. Offer’s terms of reference, agreed with the Government, were to consider whether, and if so what, changes in the electricity arrangements would best meet the needs of customers with respect to price, choice, quality, and security of supply; enable demand to be met efficiently and economically; enable costs and risks to be reduced and shared efficiently, provide transparency; respond flexibly to changing circumstances; promote competition in electricity markets, facilitating entry and exit from such markets; avoid discrimination against particular energy sources; and be compatible with Government policies (Offer, 1998d, pp. 83–84). The process that led to the eventual ending of the Pool and its replacement by NETA have been extensively described and criticized elsewhere (e.g. Newbery, 1998b, c). Shuttleworth (1999), writing after the publication of Offer’s Interim Conclusion (Offer, 1998d), noted that “it is difficult to find any rigorous analysis to underpin the reform proposals”, while Newbery (1998c) concluded that “(T)he present review appears to have relied mainly upon

17

That is, Eastern paid £6/MWh to the selling company for all electricity generated, increasing the effective marginal cost by that amount.

Electricity Liberalization in Britain and the Evolution of Market Design

127

unsubstantiated claims, inappropriate analogies, unquantified criticisms, and a remarkably uncritical assessment by the participants of the debate, without commissioning the kind of detailed analysis one might have expected from a regulatory agency claiming industry expertise.” The Pool Review (Offer, 1998e) argued that the complexities of price formation in the Pool allowed generators to exercise more market power than would have been possible had the market been structured more like a classic commodity market. It criticized the opaque method of determining price based on a scheduling algorithm (GOAL) devised for the vertically integrated CEGB, as well as the capacity payments, and the concept of a single-price auction. It also criticized the PSA for blocking desirable changes, because as a contract between parties it could only be changed with their agreement, and, given the voting arrangements, it was rare for any change to make all parties better off. The recommendations of the Pool Review were accepted after extensive industry consultations and NETA went live on March 27, 2001. NETA replaced the PSA by a Balancing and Settlement Code with a more effective method of making modifications, giving Ofgem more influence in the process.18 The Pool ceased to exist. Electricity was now to be traded in four voluntary, overlapping and interdependent markets operating over different time scales. Bilateral contract markets cover the medium and long run, while forward markets offer standard contracts (base-load, peak hours) for periods up to several years ahead. A short-term “prompt” bilateral market (OTC and exchange), operating from at least 24 hours to Gate Closure (described below, initially 31⁄2 hours before a trading period, subsequently reduced to 1 hour in July 2002), allowed parties to adjust their portfolio of contracts to match their predicted physical positions. This short-term market would yield information to construct a spot price for each half-hour (e.g. the UKPX Reference Price Data). At Gate Closure, the official end of the bilateral markets, all parties had to announce their Final Physical Notifications (FPNs) to the SO. The SO would then accept bids and offers for balancing the system. These bids and offers would be fed into the Balancing Mechanism (BM) to produce cash-out prices for clearing imbalances between traders’ FPNs and their actual (metered) positions. The most obvious difference between NETA and the Pool is that under the Pool all generation was centrally dispatched while under NETA plant is self-dispatched. The obligation to balance output with demand is now placed on each generator, with the SO’s task confined to ensuring system stability. The Pool, that previously acted as both a wholesale market for all electricity and allowed NGC as SO to balance the system, was replaced by a bilateral market and a BM (also operated by NGC as SO) for the residual imbalances of parties that fail to self-balance. Whereas the Pool operated as a uniform single-price auction for buying and selling all power (including that needed for system balance), under NETA the market participants themselves determine the prices for the great bulk (about 97%) of all power traded. The BM is run as a discriminatory (pay-as-bid) auction for the residual amount (about 3%). Participants pay the cost of remedying any imbalances in their own positions, and NGC charges for system balancing through the Balancing Services Use of System charge. Elexon determines two cash-out prices: the weighted average of accepted offers determines the system buy price (SBP) and that of bids the system sell price (SSP). Any party found to be out-of-balance when metered amounts are compared with FPNs is charged either the SBP (if they are short, that is the FPN is more than the metered output (for a

18

Ofgem, the Office of Gas and Electricity Markets, was set up in 1999 as the successor to Offer.

128

Electricity Market Reform

generator) or less than metered consumption (for a consumer), or they receive the SSP if they are long (and have to spill power). The critical feature of the original design of the BM is that these prices are normally different (SBP ⭓ SSP),19 and charge each party’s imbalances, regardless of whether or not they amplify or reduce the system imbalance as a whole. Figure 4.8 below gives an indication of this volatility. As a result of the initially extreme volatility of the balancing prices and the high values taken by the SBP, a considerable number of modifications were made. One of the more important modifications (P78, shown in Fig. 4.6) made the reverse balancing price (i.e. the price facing parties who were in the opposite position to the overall market, e.g. long when the market was short, and hence aiding balance) would revert to the spot price, and hence not penalize those helping balance the system relative to their selling in the spot market. The ideas of moving to a single balancing price, or a balancing price based on marginal rather than average cost, have has been mooted but so far rejected by Ofgem. Note that there are two distinguishing characteristics of the BM, either of which could be changed independently. The first is that there are (normally) two different prices for being short or long. The second characteristic is that these prices are determined from a discriminatory auction in which bids and offers pay or are paid as bid, and the average cost of securing the services is then charged out.20 One consequence of this combination is that it is more risky for a generator to offer balancing services. If a generator has an accepted offer to increase output, and then suffers a loss of output, he is likely to have to pay more than he is paid. He may therefore prefer to retain the spinning reserve for his own insurance. A single final balancing price would make such an offer never any worse than self-insuring and normally better, and would thus promote a more liquid balancing market.

4.4.1. The evolution of market structure During the period in which NETA was under discussion, National Power and PowerGen had to decide on their future strategy in the face of considerable uncertainty about market developments and impending excess capacity. Until 1995, the RECs had been protected against take-over by Golden shares, but these lapsed and in the following few months eight of the 12 RECs were targeted by bidders. Six were successfully acquired, two by other UK regulated utilities, one by the vertically integrated Scottish electric utility, one by Scottish Power, and two by US utilities. The bids by National Power and PowerGen for two RECs were referred to the Monopolies and Mergers Commission, and then blocked by the Secretary of State. One of these RECs was subsequently bought by another US utility group. The logic of combining risky generation with the offsetting risks of supplying downstream customers was amplified by the very favorable low-debt position of these regulated utilities, and made them irresistible to the duopoly generators, but the market power of the

19

The prices were equal by about 25% of the time, and SSP exceeded SBP very occasionally (0.1% of the time) in the first 18 months. 20 The Dutch balancing market is at the other extreme. It operates a uniform price auction to determine a single price for those 15-minute periods in which the system is either long or short for the whole period, and charges those who are short while rewarding those long. There is the potential (not yet used) to add a penalty of 1 euro/MWh to both imbalances. If the system is both short and long within the 15-minute period it determines two prices, effectively one for each sub-period in which the imbalance is in one direction.

Electricity Liberalization in Britain and the Evolution of Market Design

129

duopoly generators had already led the Secretary of State to reject their proposed vertical integration on one occasion.21 The obvious solution was for the companies to divest generation so that they could pass scrutiny when they bid for supply companies, and that in preparation for full retail liberalization were being unbundled from the REC distribution businesses. The urgency of achieving this objective was increased by the uncertainty over the new trading arrangements. Thus on November 25, 1998 PowerGen entered undertakings with the Secretary of State to sell 4000 MW of plant and to end the earn-out clause on its 1996 power station sales in return for clearance to acquire East Midlands Electricity’s distribution and supply business. Similarly National Power agreed to sell the 4000-MW Drax station in order to buy the supply business of Midlands Electric. The delicate task facing National Power and PowerGen was to sell the plant for attractive prices into a market that was in danger of being oversupplied with increasing gas generation. Here the new Labour Government may have inadvertently helped by imposing a moratorium on building new gas-fired plant in 1997 to assist the coal mining industry during the period of sorting out the Pool. (The Government also imposed a so-called Climate Change Levy that was actually a tax on energy rather than carbon, again protecting coal, if not British Coal.) The DTI estimated that the moratorium delayed the building of 5800 MW of gas-fired capacity (some indefinitely). The solution for the duopolists was to ensure that the price–cost margin remained high while plant was offered for sale, and Sweeting (2001) identifies the period from 1998 to early 2000 as one in which National Power and PowerGen could have increased their individual profits if they had bid lower prices. That is consistent with co-ordinating on a higher-price equilibrium than short-run myopic profit maximization would deliver. During this period plant was profitably sold, indicated by the falling HHI in Figure 4.5, and the changing shares in Figure 4.7. The companies buying the plant were warned by Ofgem that there was no guarantee that prices would remain high, particularly given the impending arrival of NETA, which Ofgem was claiming would itself lead to prices at least 10% lower than otherwise. Nevertheless, Edison Mission paid £1.3 billion for the 2000-MW stations at Fiddler’s Ferry and Ferrybridge in July 1999, or £314/kW, and increased the plant output by more than 30%. With the new buyers keen to improve the returns on their purchases by increasing plant output, Figure 4.5 shows that the earlier co-ordinated duopoly equilibrium was no longer sustainable and the price–cost margin collapsed before NETA went live, but after the fall in concentration (HHI). Edison Mission subsequently sold its two stations in October 2001 for less than half the purchase price (incurring a balance sheet impairment of $1.15 billion on the $2 billion purchase cost). Plant sales have continued apace, and DTI (2004) notes that during the Summer of 2004 a number of US firms, including AEP and Edison Mission Energy, exited, while major electricity companies (SSE, Scottish Power, Centrica) bought stations from US owners or their creditors. Nearly 8000 MW of generating capacity, around 10% of GB generating capacity, changed ownership, with the majority going to vertically integrated companies with interests in both electricity generation and supply.

21

Risks could have been hedged by long-term contracts between generation and supply companies, but the transaction costs of writing long-term contracts to cover all contingencies (such as the ending of the Pool, the Emissions Trading System, Climate Change Levy, Renewables Obligation Certificates) might make vertical integration more attractive. Supply companies also suffer from credit risk as they are typically under-capitalized unless combined with generation or distribution.

130

Electricity Market Reform 40,000

Offer encourages sale of 6000 MW

NETA live

35,000 PG and NP trade horizontal for vertical integration

30,000

MW

25,000 20,000 15,000 10,000 5000

April 1990 October 1990 April 1991 October 1991 April 1992 October 1992 April 1993 October 1993 April 1994 October 1994 April 1995 October 1995 April 1996 October 1996 April 1997 October 1997 April 1998 October 1998 April 1999 October 1999 April 2000 October 2000 April 2001 October 2001 April 2002 October 2002 April 2003 October 2003 April 2004 October 2004

0

ALCAN Innogy International Power

National Power (NP) Edf TXU/Eastern

British Energy Independent AES AEP Edison Powergen (PG) SS&E

Fig. 4.7. Capacity ownership of coal generation, 1990–2005. Source: Data supplied by Bower and Humphries.

4.4.2. The impact of the new trading arrangements on market performance The intellectual case for replacing the Pool through which all energy was traded by a voluntary BM covering rather less than 3% of energy was that this would force both buyers and sellers to haggle over the price of electricity without the clear and transparent signals delivered by the Pool. If parties were forced by a risky, opaque, and potentially penal imbalance market to contract ahead, consumers would have time to shop around for better deals, making the market more competitive. This argument ignores two important facts. The first is that about 90% of electricity traded before NETA was under contract, and the annual contract round was a period of intense haggling. NETA did not change that. The second is that the relative bargaining strength of generators and consumers depends more on the ability of generators to hold the market to ransom than the fine details of the spot or balancing market. This strength is best measured by the extent to which consumers can meet their total demand from all other generators, and the number of generators that are pivotal (i.e. essential) for meeting demand. The loss to a generator of not selling is the difference between the price and variable cost (shown in Fig. 4.5), whereas that to a consumer of not being able to buy power is potentially the difference between the VOLL and

Electricity Liberalization in Britain and the Evolution of Market Design

131

the price, which may be hundreds of times as large. Pivotal generators therefore have very considerable bargaining power in any market design. Three factors influence this bargaining power: the number of competing generators, the reserve margin, and (in the longer run) the ease of entry. Entry was extremely easy in the Pool, but, after vertical integration and with the removal of a guaranteed market of final resort, considerably riskier under NETA. The Pool reserve margin was normally quite adequate as a result of earlier entry, while competition had just become intense as the Pool ended, and was already demonstrating its impact on spot and contract prices. The claim that NETA therefore was necessary (and sufficient) to mitigate generator market power is unsubstantiated. There was a rather more confused claim that replacing a single-price auction (like the Pool) by a pay-as-bid or discriminatory auction would obviously lower the average price, ignoring the auction literature on revenue equivalence. A more sophisticated claim was advanced by Currie (2000), who argued that repeated single-price auctions encouraged collusion more than discriminatory auctions. Newbery and McDaniel (2003) argued that the theoretical, empirical, and experimental evidence on auction design applied to the electricity market was ambiguous. Fabra et al. (2004) developed simple models comparing the two auction designs and were able to demonstrate that with predictable and unchanging demand, a discriminatory auction would yield lower (short-run) prices than a single-price auction, but would typically lead to a less efficient use of plant. They were not able to produce results for multi-period and repeated auctions. Offer (1999) estimated the costs of switching to NETA at about £700 million (spread over a 5-year period) followed by additional annual costs of £30 million.22 Offer justified this cost by claiming that prices would fall 10% as a direct result (although this would be a transfer from generators to consumers, not a net social benefit). Removing the gas moratorium helped, but in the event prices fell so far that CCGT entry was put on hold indefinitely. Ofgem was able to claim that the “Evidence of the first year of NETA shows wholesale prices around 40% below those under the former Electricity Pool.” (Ofgem, 2002). While this may exaggerate the fall, Figure 4.5 shows that compared to 1999, prices at the time of NETA were indeed substantially lower. Newbery and McDaniel (2003) argued that the price fall was due to competition, not NETA, as prices fell before NETA, and as electricity cannot be stored, future prospects of changed trading should have had no impact on pre-NETA prices. Bower (2002) (who supplied the plant data used in Fig. 4.6) demonstrated this more rigorously using formal econometric tests. Evans and Green (2003) confirmed this finding for Bower’s specification, but raised the question whether the announced end of the Pool would unravel the collusive equilibrium by backward induction for the point at which the Pool (and transparent pricing) would end. They found support for a variable “para-NETA” that takes the value 0 until October 2000, and 1 thereafter, using data up until September 2002. Prices and margins have subsequently risen, and these regressions do not perform as well on the longer time series now available.23 An alternative and simpler explanation is that indeed collusion ended, not by the anticipated arrival of NETA, but because of the actual fall in concentration, an event carried forward by the desire of the incumbents to integrate forward into supply, expressed as early as 1995. 22

“The costs of implementing and operating the new trading arrangements are estimated to be between about £136 to £146 m per annum, for a 5-year period. Thereafter the operating costs are expected to be of the order of £30 m per annum.” (Offer, 1999, p. 14). These continuing costs almost certainly understate the extra costs of maintaining 24–27 trading floors for balancing. 23 Personal communication regarding work in progress from Green.

132

Electricity Market Reform

4.4.3. Security of supply in an energy-only market The Pool paid increasing amounts for capacity as the reserve margin declined, to encourage generators to maintain an adequate reserve margin, either by keeping old plant open longer or by investing in new plant. Its determination was described above and implements the solution proposed by Vardi et al. (1977), itself the result of a long debate on peak-load pricing dating back to Steiner (1957), Turvey (1968), and Boiteux (1960). NETA abandoned this principle, and left generators to ensure that they contracted and bid to cover not just energy costs but also capacity costs. The central question faced by market designers is whether a liberalized competitive energy-only electricity will deliver adequate security of supply, or whether some additional mechanism, such as a capacity payment or capacity obligation, is required to support an adequate reserve margin for generation adequacy and system reliability. In the medium run the average price in any electricity market will be determined by the conditions and costs of entry and exit (and possibly on the threat of price caps or other regulatory interventions that might reduce expected profits). The first publication of the Joint Energy Security of Supply (JESS) Working Group in June 2002 (DTI, 2002) stated that “Capacity margins are healthy and are expected to remain so.” Their next report in February 2003 indicated no change. Shortly thereafter, generating companies started to experience financial distress and some went into administration. British Energy, the privatized nuclear company, had to be bailed out by the Government, and surviving companies started to scrap or mothball plant. The plant’s reserve margin fell to below the NGC’s target margin of 20%, and in the Summer of 2003, NGC’s forecasts suggested a rapidly deteriorating situation. In response, Winter 2003/2004 forward peak prices rose from £25/MWh to over £35/MWh and some mothballed plant was returned to the system. Some of the Winter month forward prices rose even more dramatically as Figure 4.8 shows. In the event the Winter was mild (only 15% of Winters in the past 75 years were as mild), demand was lower than the previous year, and there were no capacity shortages. The Third JESS Report in November 2003 (DTI, 2003) claimed that forward markets were delivering the appropriate signals and participants were responding as they should, although the scare revealed a worrying lack of information about the status and likely time needed to return mothballed plant. The impact of plant removal and narrowing reserve margins can be seen in Figure 4.5, where the price–cost margin has returned a considerable way toward a more sustainable equilibrium (although one that in 2004 was still too low to justify new build). The Fifth JESS Report of November 2004 (DTI, 2004) forecast the plant margin for Winter 2004/ 2005 at 20%, and noted that there was a further 1.2 GW of mothballed plant that could return within this Winter period if required, which would raise the plant margin to over 22%. Ofgem has argued that markets worked well and signaled impending scarcity in adequate time for the system to respond by returning plant that had been withdrawn primarily in response to the collapse in price as the markets equilibrated to their newly competitive structure. Sceptics argue that until NGC announced the impending scarcity little happened, and that 6 months’ warning is inadequate for delivering anything other than recently mothballed plant (as the longer they are idle, the more likely it is that they will be cannibalized for spares and will take longer to return to service).24 In addition, following pressures from Ofgem and analysis of its likely reserve requirements for Winter 2003/2004, NGC initiated 24

Britain may be relatively unusual in the quantity of mothballed plant that has been potentially available for return to the system. In other jurisdictions the opportunity cost of maintaining such plant, rather than securing tax deductions for abandoning plant and releasing the site for replacement construction, may militate against mothballing.

Electricity Liberalization in Britain and the Evolution of Market Design

133

60 Base load November 2003 Base load January 2004 Peak November 2003 Peak January 2004

Forward price (£/MWh)

55 50 45 40 35 30 25

03 gu st 20 27 03 Au gu 16 st 20 Se 03 pt em be r2 06 00 O 3 ct ob er 20 26 03 O ct ob er 15 20 N ov 03 em be 05 r2 D 00 ec 3 em be 25 r2 D 00 ec 3 em be r2 00 3

20

Au

ly

07

Ju 18

28

Ju

ne

20

03

20

Fig. 4.8. Winter 2003/2004 forward electricity prices (£/MWh). Source: Heren data.

a Supplemental Standing Reserve Tender on October 14, 2003 to increase its reserve capacity. The tender closed on October 27, 2003, and NGT procured a total of 852 MW of Supplemental Standing Reserve at a total cost of £18.87 million or £22/kW (Ofgem, 2004a). The majority of this volume was provided from plant that had previously been mothballed and a significant amount came from the demand side. What can we conclude from this expensive change in market arrangements and to what extent has security of supply been affected? First, the costs of balancing are now directed on those that cause imbalances, and to that extent should induce more efficient responses. On the other hand, these balancing charges may now be excessive, to the detriment of nonportfolio generators (i.e. new entrants and British Energy) and intermittent suppliers like wind.25 The balancing prices are considerably more volatile and unpredictable than the pool prices that served as a more liquid balancing market. Figure 4.9 shows monthly moving averages of the buy and sell prices as well as the spot prices. The weekly or daily prices show considerably greater volatility, and a measure of this is provided in Figure 4.10, which gives the standard deviation of the difference of the spot and balancing prices for a period of a week (i.e. for the 336 observations) and a quarter (i.e. 4368 observations). The effect of the P78 rule change is very clear where the difference between the SBP and SSP falls sharply in March 2003 and remains moderately low thereafter. The impact on volatility is somewhat less pronounced but statistically significant. 25

Although the net surplus of the BM is recycled, there are transfers between different types of participants, while there are extra real costs in the dual cash-out prices which fail to confront participants with the correct costs of centrally provided balancing, especially in maintaining extra spinning reserve.

134

Electricity Market Reform 50 SBP UKPX RPD SSP SBP⫺SSP

45 40

P78 rule change

£/MWh

35 30 25 20 15 10 5

January 2002 March 2002 May 2002 July 2002 September 2002 November 2002 January 2003 March 2003 May 2003 July 2003 September 2003 November 2003 January 2004 March 2004 May 2004 July 2004 September 2004 November 2004 January 2005 March 2005 May 2005

0

Fig. 4.9. Spot and cash-out monthly moving average prices January 2002 to June 2005. Source: Elexon price data: UKPX is the Reference Price Data for the day-ahead spot market. 90 Monthly SD of SBP ⫺ RPD

80

Quarterly SD of SBP ⫺ RPD Quarterly SD of RPD ⫺ SSP

70

Incident at Damhead Creek SSP goes to ⫺£5870

£/MWh

60 50

P78 rule change

40 30 20 10 1 May 2005

1 March 2005

1 January 2005

1 November 2004

1 July 2004

1 September 2004

1 May 2004

1 March 2004

1 January 2004

1 November 2003

1 July 2003

1 September 2003

1 May 2003

1 March 2003

1 January 2003

1 November 2002

1 September 2002

1 July 2002

1 May 2002

1 March 2002

1 January 2002

0

Fig. 4.10. Standard deviation of difference of balancing price and spot price 2002–2005. Source: Elexon price data: UKPX is the Reference Price Data for the day-ahead spot market.26 26

On May 19, 2004 NGC engineers took action to prevent an unsafe situation and issued its first ever emergency instruction to stop Damhead Creek Power Station. This in turn triggered the acceptance of

135

Electricity Liberalization in Britain and the Evolution of Market Design

110 NETA BM 03-4

BM 03-4 Pool 1998–1998

35

90

Pool 1997–1998

70

30

50

25 6

0

4.

5

4.

9

3.

4

2.

8

2.

3

1.

7

1.

2

0.

20

0.

0

30 0.

Cost net of avoidable cost £/MWh

40

Percent

15 10 5 0

0

10

20

30

40

50

60

70

80

90

100

Percent time cost higher than marginal cost by amount on y-axis Fig. 4.11. Cost of 24-hour failure under the Pool and NETA. Source: Pool data and UKPX RPD data.

The costs of balancing will depend on whether the participant is a generator or supplier. As contestability is a key issue, the relevant question is whether the BM unreasonably raises the risk to a small entrant. This can be estimated as the risk of having to pay the buy price (SBP) after a generator suffers a forced outage, in order to meet an assumed contract position. If a generator fails at a random moment and stays off-line for 24 hours, the cost will be the 24-hour average of the SBP from that moment. In the year before the P78 rule change indicated above, the expected cost of such an outage (relative to an assumed variable cost of £12/MWh) was £17/MWh or £0.4/kW/event compared to £13/MWh or £0.32/kW/event under the Pool for 1997–1998. The variance was, however, twice as high as under the Pool. In the year following P78, the average cost had fallen to £11/MWh or £0.3/kW/event and the variance had also fallen to 150% that of the Pool. Figure 4.11 illustrates the cost duration curve for balancing under NETA from April 1, 2003 to March 31, 2004 compared to the Pool in 1997–1998. Thus 5% of the time the cost would be £30/MWh for the following 24 hours in both the Pool and the recent BM, and 1% of the time it would be £70/MWh in the BM compared with £44/MWh under the Pool. The risks in the early days of NETA were very much higher and led to claims that plant was inefficiently part loaded to avoid penal imbalance costs, at considerably higher cost. One should interpret this finding with some care, as the Pool required bids to remain valid for 24 hours while bids and offers to the BM can be changed on a short time scale and in response to a perceived tightening of the market when a large unit goes off-line, making it more risky for generators to handle outages. Even if we ignore such responses, if a large plant were to go down, the demand in the BM might be such as to considerably increase the 26

(continued) a large proportion of bids at ⫺£9999/MWh, causing the SSP to go to ⫺£5,728 in HH 28 at a cost of £3.55 million (Elexon Circular EL01201 of August 20, 2004). This had a huge impact on the standard deviation of prices for the whole of the 30-day window of the moving average, swamping all other prices, as can be seen.

136

Electricity Market Reform

short-run cost, but without knowing the shape of the bids and offers it is hard to estimate by how much. The effect on suppliers is that they would over-contract on average as the SBP is more costly than spilling the surplus at the SSP, and this would (slightly) raise the cost and risk of selling. This may have the desirable effect of encouraging contracting, which tends to mitigate generator market power in the spot market, although at the expense of increasing demand and hence market power in the contract market. The low liquidity of most electricity markets (and certainly the British markets) makes the cost of rebalancing contract positions high and again acts as an entry barrier. Second, the BM mutes scarcity signals by paying generators their bid price and not the marginal price (in order to mitigate market power and reduce possibly politically unacceptable price spikes). This, together with the lack of integration with spot, forward and contract markets, and the lack of any capacity payment, makes the entry decision more uncertain and risky, and may lead to lower reserve margins. If so, then the lower reserve margins and the extra entry costs will allow a higher-average wholesale price, refuting Ofgem’s claim that NETA alone (i.e. regardless of market structure) would reduce wholesale prices by 10%. Again, this claim needs to be examined carefully, for example, for a peaking generator that can offer balancing services of a half-hour duration at very short notice. Figure 4.12 shows the annual profits that a peaking unit (with fast start-up able to offer half-hourly supplies and selling into the balancing market at the SBP) would earn at different prices for its variable costs. As an indication, the June 2005 cost of distillate at 32% efficiency would have been about £75/MWh, ignoring the (high) fuel excise, but less than half that running on natural gas. One obvious problem with this calculation is that it requires skilled bidding to achieve the SBP in a pay-as-bid market. The graph demonstrates that profits are sensitive to the fuel cost and balancing prices (which are likely to be related). For example, taking the daily British gas prices (spot at the National Balancing Point), the profit of a 32% efficient opens-cycle gas turbine in 2002/2003 would have been £248/kW/year. The following year when gas prices rose and balancing prices were less volatile, the profit would only have been £21/kW/year, but in 2004/2005 the profit would again have risen to £145/kW/year. The annual fixed costs would include grid charges (which vary across the

200 2002/2003 2004/2005

180 160 £/kW/year

140 120 100 80 60 40 20 0 20

30

40

50

60

70

80

90

100

110

120

Variable cost £/MWh Fig. 4.12. Profit of a peaking plant selling into the BM at SBP. Source: Elexon price data.

Electricity Liberalization in Britain and the Evolution of Market Design

137

country) from ⫹£9/kW/year to ⫺£7/kW/year) and other fixed costs (perhaps £6/kW/year), as well as capital charges of perhaps £36/year, suggesting a requirement of £50/kW/year. This suggests that peaking investment can be remunerative under NETA, but is risky, partly because of fuel and electricity price uncertainty, and partly because of the problem of bidding. This could be circumvented by offering such services to NGC on contract to bid into the BM, and/or making the BM a single marginal price market. NGC simulated the effect of moving to a marginal price balancing market for 2003/2004, although ignoring the likely response of participants to such a significant change in market rules (NGT, 2004). The average SBP would have increased by £2.78/MWh and SSP decreased by £1.80/MWh. The standard deviation of SBP would have increased by £14.39/MWh indicating an increase in volatility, whilst the increase in standard deviation of SSP was much smaller at £1.23/MWh.27 This increased volatility would have improved the profits of peaking plant (while perhaps increasing its risk somewhat). Critics of energy-only markets argue that the new system has not been properly stress tested yet, and that reliability either requires a sufficiently concentrated market to ensure high enough prices to cover full costs, or some supplementary mechanism to reward capacity and encourage investment in a timely manner in advance of possible system scarcity. Thus the US Federal Energy Regulatory Commission (FERC) in its 2002 Notice of Proposed Rule Making on the Standard Market Design stated: “FERC is concerned that the spot market alone will not signal the need to bring development of new supply resources in time to avert a shortage … .” The British approach is to argue that provided markets are allowed to clear and are not subject to an unreasonable price cap,28 and provided the TSO is incentivized to maintain system reliability, a competitive energy-only market should work well without additional mechanisms.29 The argument is that the spot and balancing market should signal short-run scarcity and force consumers to contract ahead, thus feeding scarcity signals back into forward and contract markets. As already argued, this would be better achieved by a marginal price balancing market than the British design of an average price of the offers as paid. The main concern is whether investors will look far enough ahead to make timely investments. In a tight oligopoly, competition for market share and/or high entry-inducing prices should provide adequate incentives. If the market is as competitive as in Britain then market risk is more likely to lead to delays in exercising the real option of investment as the preemption effect will be lower. Some (e.g. de Vries and Hakvoort, 2002; de Vries and Neuhoff, 2004) argue that generation investment requires long-term contracts like the earlier PPAs in the Pool, and/or a captive franchise market on which to write them. Others (e.g. de Luze, 2003) are concerned that generating companies’ recent financial distress (and banks who might provide debt) will be reluctant to invest until profits significantly improve.

27

NGT simulation’s results should be interpreted with care as they rely on the strong assumption that generators and suppliers would not have changed their bidding behavior into the market; there is a large literature demonstrating the impact of market rules on market players bidding strategies. 28 It is reasonable to cap spot markets by the VOLL, and the BM in Britain can only enter four digits, making the maximum allowed price £9999. 29 Part of the problem in the USA is that the ISOs are prone to take “reliability actions”, including out-ofmarket calls, as supplies get tight that have the effect of depressing prices. It is admittedly difficult to ensure that balancing prices (and local spot prices) efficiently reflect scarcity and not market power. There is also the different institutional history of managing privatized electricity utilities in Britain and the US that may partly explain differences in the confidence which regulators place in market solutions to generation adequacy.

138

Electricity Market Reform

However, most generating investment in Britain is likely to be undertaken by the large Continental companies already present – EdF, RWE, and E.On, who are also vertically integrated into supply and hence well hedged in the wholesale market. Such companies are unlikely to rely on short- to medium-run forward contracts in making investment decisions that last 30–40 years, and will base their decision on market fundamentals. Although the British market is competitive, these players have ambitions to retain their Continental position with high levels of output, and seem likely to invest if the fundamentals warrant. The early complaints of wind generators and combined heat and power (CHP) that they were discriminated against have not been adequately tested against more recent market conditions – certainly CHP output dropped dramatically but that was arguably because of the adverse spark-spread (electricity less gas cost). Wind power now typically sells on contract to supply companies who can better manage the imbalance risk within their entire portfolio, and is in any case massively rewarded by Renewable Obligation Certificates that increase the electricity price from around £25/MWh to £65/MWh. Perhaps the greater concern is that such a high level of renewables support may not be credible in the longer run, and this policy uncertainty may make investment in all forms of generation riskier.

4.5. Retail Competition Full retail competition was envisaged right from the start of restructuring, with a phased opening, initially for the 5000 or so customers of above 1-MW demand, then in 1994 for the 50,000 sites with above 100-kW demand, and finally the aim was complete liberalization by 1998. By offering customers to freedom to switch supplier, suppliers would be put under pressure to cut margins, offer tariffs and contracts suited to the customer, and in turn bid more aggressively in the wholesale market, and the ability of the regulator or government to influence contracting and generation decisions would be reduced. Large customers mostly switched from their local RECs to the two incumbent fossil generators, National Power or PowerGen. The size of the competitive market increased in April 1994, when the 50,000 sites with demands of between 100 kW and 1 MW were allowed to change their supplier, and required to install a half-hourly meter. The RECs were initially slow to enter the second-tier market to supply large customers,30 but thought they could (and would have to) compete more effectively for smaller customers. One-quarter of the RECs entered the 100-kW second-tier market in the first year, and the share increased steadily thereafter. By 1996, they (collectively) supplied more than half the sites taking a competitive supply, with greater success in the 100-kW market. Offer’s early survey of suppliers indicated that competition had been effective even among the smallest sites in this market, those with demands of under 300 kW, for one-quarter of them bought from a second-tier supplier in 1995/1996. By 2002, 62% of half-hourly metered customers had switched at least once (Ofgem, 2003, Table 5.1). Price controls in this market were lifted in 1994, when competition was introduced, and the ability to change supplier seems to have been sufficient protection for these consumers. If competition was successfully introduced, the process itself was fraught with difficulty, as significant changes to systems and procedures were needed to cope with the larger number of customers (Green and Newbery, 1997). Many important decisions were not made 30

First-tier customers buy from the local (originally franchise-holding) REC, second-tier customers buy from other suppliers, so a REC selling out of area is a second-tier supplier. This distinction was removed in the Utilities Act of 2000.

Electricity Liberalization in Britain and the Evolution of Market Design

139

early enough to allow systems to be developed and tested. Installing the required meters and communications link was chaotic, and many were either not installed in time, or installed but not properly registered. It took more than a year to sort out some of the problems, amid acrimonious debates over who had caused the chaos.31 Since then competition appears to be developing well, measured by periodic surveys published on the Ofgem web site (e.g. Ofgem, 2003), despite recent consolidation and some increased concentration in the supply market (with an HHI of 1670 and the top six suppliers having 90% of the market at the end of 2002). Ofgem’s (2000) review noted concerns about potential entry barriers caused by a lack of unbundling of distribution and supply, but legal unbundling was subsequently required by an amendment in the Utilities Act 2000. Full retail liberalization started in September 1998 and was completed by June 1999, with a near-90% consumer awareness of the opportunity to choose (Littlechild, 2001). Ofgem (2004b) reviewed the experience, noting that about half of domestic customers had switched at least once (of whom 80% moved to a supplier with a “dual fuel” offer). British Gas, the original gas franchise holder, had 44% of the dual fuel market despite being the most expensive (for gas). Incumbents maintained prices in their original franchise area substantially higher than they offered to new customers out of area, and genuinely new entrants (those without an existing franchise business) collectively secured less than 1% of the market, most of them going bankrupt or selling out to incumbents. The incumbent Eastern/TXU also went bankrupt and many RECs sold their supply businesses. Prices fell, but not as much as costs, with incumbent supply margins widening substantially (up to 26% of domestic bills), first under generous price caps on within-area retail electricity designed not to stifle competition, and subsequently, when price controls were removed, because of the stickiness of customers. Salies and Waddams Price (2004) estimated that the incumbency effect raised prices 4–13%. By 2002 a domestic customer cost about as much on the stock exchange as 2 kW of generation (rather more than the capacity needed to supply him/her). Ofgem (2004b) noted that the market was “competitive but not yet mature”, and by comparison with most other examples of electricity liberalization, Britain has been judged a success, notably as a high proportion switched suppliers. Nevertheless, Green and McDaniel (1998) noted that retail liberalization had cost £276 million to set up with on-going costs estimated at £36 million/year, while Salies and Waddams Price (2004) concluded that overall the net social benefits of liberalization were negative. Taking a broader view, retail liberalization allows the regulator to stop interfering in this part of the market, and transfers the risks of contracting to suppliers, who can no longer pass through “sweetheart” deals to a captive supply base. Ministers and politicians would find it harder to protect the coal industry with coal-backed electricity contracts passed through to the franchise market, as in 1993–1998, and the ability of the Government to protect renewables at the expense of consumers has required an Act of Parliament. Pessimists (de Vries and Neuhoff, 2004) have argued that the resulting lack of long-term contracts backed by a captive market makes entry into generation more difficult, and might reduce security of supply. On the other hand, the high profits in retailing have compensated 31

Blame was heaped on the electricity companies, for delaying decisions; the regulator, for introducing competition in metering services at the same time; and the meter operators, for failing to install meters when promised (even those not wishing to change supplier required a half-hourly meter and communications equipment). None of the parties faced financial penalties for failure to meet their obligations (although this was apparently considered by Offer), and the industry recovered part of the additional cost of sorting out the disruptions from a levy on consumers.

140

Electricity Market Reform

to some extent for the losses caused by the margin collapse in the wholesale market (and vice versa when wholesale prices rise), and vertically integrated companies are effectively automatically contracted when customers are as sticky as domestic households. Vertical integration, provided it is combined with adequate competition, may provide the financial stability and market assurance needed to ensure adequate investment in generation. Most domestic customers would probably be better off with a regulated supply margin and benchmarked contract costs passed through under regulatory supervision. In short, it is difficult to see why retail liberalization attracts so much support in the new Electricity Directive that compels full market opening by 2007 (other than the reasonable assumption in some jurisdictions that regulators would be less effective at protecting consumer interests than competition). 4.6. Conclusions The British experiments have demonstrated a number of important lessons for electricity market liberalization. First, ownership unbundling of transmission from generation helps support a competitive wholesale market, which in turn puts pressure on companies to reduce costs. Scotland failed this test and failed to improve its performance. Second, efficient pricing of scarce interconnector capacity and charging correctly for losses might have allowed the Scottish market to import English competition, but was blocked by the courts. Third, while competition drives down costs, concentrated markets can sustain inefficiently high price–cost margins. Pivotal generators retain market power that is best addressed by reducing concentration, although entry that increases the reserve margin also helps. Tacit co-ordination is likely given electricity market characteristics, and is best addressed by encouraging contracts and entry, and reducing concentration. Fourth, investment in generation can be facilitated by a transparent Pool, domestic franchises, and wholesale market power. Britain replaced nearly one-third of its (already adequate, if rapidly obsolescing) capacity by such means. Entry (including returning mothballed plant to service) is responsive to price signals (the forward spark-spread). Fifth, unbundling and liberalization increase risk for generators and encourage them to seek vertical integration with suppliers. This offers the opportunity for the regulator and competition authorities to trade horizontal for vertical integration and to reduce concentration, at the cost of increased entry barriers. A better alternative is to start from a more fragmented structure. That would allow one to consider legal restraints on such vertical integration to encourage more contracting and market liquidity, but we lack evidence on the costs and advantages of such enforced competition. Sixth, retail liberalization has delivered significant benefits to large and medium customers, although managing the market opening for more than the largest customers is challenging and can be expensive if all customers need interval meters and communications equipment. There is little evidence that full retail liberalization delivers positive net social benefits in well regulated jurisdictions, although it has enabled vertically integrated companies to hedge their wholesale market position with sticky downstream customers, enabling them to continue to finance investment when needed. Seventh, the British approach to liberalization that requires licences to be held by both potentially competitive and natural monopoly segments has worked better than many of the Continental alternatives (and arguably the USA’s onerous duty on regulators to deliver “just and reasonable” prices). Licences require the holders to provide the regulator with the information needed for adequate market monitoring, and allow market abuses to be swiftly and cheaply addressed.

Electricity Liberalization in Britain and the Evolution of Market Design

141

Finally, such apparently basic issues as the desirability or not of capacity payments (or obligations) and the design of the wholesale and balancing markets remain unresolved. NETA abolished capacity payments, and relied upon the BM to signal shortages of supply, which would feed back into spot and contract markets. The ideal of a Pool with adequate competition, capacity payments, and a better governance structure for rule changes was never tried, and might have worked as well or better than NETA, with its emphasis on bilateral contracting and dual cash-out pricing in the BM. On balance, NETA replaced the Pool’s flawed governance structure by one more susceptible to incremental improvement (though at the cost of greater regulatory uncertainty) failed to increase either the liquidity of markets or the participation of the true demand side, increased trading costs, and cost over £700 million. Once it settled down and the obvious rule changes were made, NETA probably delivers similar outcomes as the Pool from existing generation. Entry is now more difficult than before, but that is not solely due to NETA. Vertical integration has reduced the demand for suppliers to contract, the end of the domestic franchise has removed the logical counterparty to contracts with new independent generators, but the removal of the Pool as a market of last resort almost certainly raises entry costs. Just at the time that FERC has embraced the concept of a Pool (with LMP) as the benchmark for the Standard Market Design, Britain has abandoned a model whose main failing was its poor market structure and governance. Nevertheless, NETA (or its successor BETTA) may not be very different from the Pool in sustaining a reliable and secure electricity market, provided the TSO is charged to deliver that reliability, if necessary by contracting ahead and making sure that unbiassed and trustworthy information about future demand and supply is fed to the markets.

Acknowledgments I am indebted to Paul Joskow, Richard Green, Alex Henney, Karsten Neuhoff, and Fabien Roques for their help, and particularly to Stephen Littlechild for his detailed comments, even where I have not accepted them, with the usual disclaimer. This chapter is a modified version of a article originally published in the 2005 Special Issue of the Energy Journal on European Electricity Liberalization. I am indebted to the IAEE for permission to reproduce the article here, and to the UKPX for their spot price data.

References Bohn, R.E., Caramanis, M.C. and Schweppe, F.C. (1984). Optimal pricing in electrical networks over space and time. Rand Journal of Economics, 15(3), 360–376. Boiteux, M. (1960). Peak load pricing. The Journal of Business, 33(2), 157–179. Bower, J. (2002). Why did electricity prices fall in England and Wales? Market mechanism of market structure?, Oxford Institute for Energy Studies Working Paper EL O2. Currie, D. (2000). The new electricity trading arrangements in England and Wales: A review. The Beesley Lectures on Regulation. The Institute of Economic Affairs and the London Business School, 10 October 2000. de Luze, G. (2003). Investment in power generation: a banker’s perspective. Presentation at the 25th March 2003 IEA/NEA Conference in Paris. de Vries, L. and Hakvoort, R. (2002). The question of generation adequacy in liberalised electricity markets. Delft University Working Paper. de Vries, L. and Neuhoff, K. (2004). Insufficient Incentives for Investment in Electricity Generation, Cambridge, CMI EP 42, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm

142

Electricity Market Reform

Domah, P.D. and Pollitt, M.G. (2001). The restructuring and privatisation of the regional electricity companies in England and Wales: a social cost benefit analysis. Fiscal Studies, 22(1), 107–146. DTI (2002). Joint Energy Security of Supply Working Group First Report, June, available at http://www.dti. go.uk/energy/jess DTI (2003). Joint Energy Security of Supply Working Group Third Report, November, available at http:// www.dti.go.uk/energy/jess DTI (2004). Joint Energy Security of Supply Working Group Fifth Report, November, available at http://www.dti.go.uk/energy/jess DTI (2005). Digest of UK Energy Statistics 2004, available at http://www.dti.gov.uk/energy/inform/ energy_stats/electricity/index.shtml Evans, J. and Green, R. (2003). Why Did British Electricity Prices Fall After 1998?, CMI WP26, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Fabra, N., von der Fehr, N.-H. and Harbord, D. (2004). Designing Electricity Auctions, CSEM WP 122, available at http://www.ucei.berkely.edu/pubs-csemwp.html von der Fehr, N.-H. M. and Harbord, D. (1993). Spot market competition in the UK electricity industry. Economic Journal, 103, 531–546. Green, R. and McDaniel, T. (1998). Competition in electricity supply: will ‘1998’ be worth it? Fiscal Studies, 19(3), 273–293. Green, R. and Newbery, D.M. (1997). Competition in the electricity industry in England and Wales. Oxford Review of Economic Policy, 13(1), 27–46. Green, R.J. (2004a). Did English Generators Play Cournot? Capacity Withholding in the Electricity Pool, CMI WP41, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Green, R.J. (2004b). Electricity Transmission Pricing: How Much Does it Cost to Get it Wrong, CMI WP63, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Green, R.J. and Newbery, D.M. (1992). Competition in the British electricity spot market. Journal of Political Economy, 100(5), 929–953. Henney, A. (1994). A Study of the Privatisation of the Electricity Supply Industry in England and Wales. EEE Ltd, London. Henney, A. (2001). The illusory politics and imaginary economics of NETA, Power UK, March, 16–26. Hortacsu, A. and Puller, S.L. (2004). Testing Strategic Models of Firm Behavior in Restructured Electricity Markets: A Case Study of ERCOT, CSEM WP 125, available at http://www.ucei.berkely.edu/pubscsemwp.html House of Commons (1993). British Energy Policy and the Market for Coal: Report HC 237. HMSO, London. Klemperer, P.D. and Meyer, M.A. (1989). Supply function equilibria in oligopoly under uncertainty. Econometrica, 57(6), 1243–1277. Littlechild, S. (2001). Electricity regulation (Chapter 2). In P. Vass (ed.), CRI Regulatory Review 2000/ 2001. Centre for the Study of Regulated Industries, Bath, pp. 25–62. MMC (1996). National Power Plc and Southern Electric Plc: A Report on the Proposed Merger. Monopolies and Mergers Commission Cm 3230. HMSO, London. Newbery, D.M. (1992). Supply function equilibria, mimeo, Cambridge University. Newbery, D.M. (1995). Power markets and market power. Energy Journal, 16(3), 41–66. Newbery, D.M. (1998a). Competition, contracts and entry in the electricity spot market. RAND Journal of Economics, 29(4), 726–749. Newbery, D.M. (1998b). Pool reform and competition in electricity (Chapter 5). In M. Beesley (ed.), Regulating Utilities: Understanding the Issues. Institute of Economic Affairs, London, pp. 117–166. Newbery, D.M. (1998c). The regulator’s review of the English Electricity Pool, Utilities Policy, 7(3), 129–141. Newbery, D.M. and McDaniel, T. (2003). Auctions and trading in energy markets – an economic analysis (Chapter 10). In P. Vass (ed.), CRI Regulatory Review2002/ 2003. Centre for the Study of Regulated Industries, Bath, pp. 195-–234. Newbery, D.M. and Pollitt, M.G. (1997). The restructuring and privatisation of the CEGB – was it worth it. Journal of Industrial Economics, XLV(3), 269–303. NGC (1997). Report on the Tender for the 1997/98 Standing Reserve Service. National Grid Company plc.

Electricity Liberalization in Britain and the Evolution of Market Design

143

NGC (2004). National Grid Company Consultation on Options for Allocating GB Transmission Access Rights under BETTA, 21 May 2004, Warwick. Available at http://www.ofgem.gov.uk/ NGT (2004). Response to OFGEM Consultation, Electricity and Gas Cashout Review, Executive Summary. National Grid Transco, Warwick. Offer (1992). Review of Pool Prices. Office of Electricity Regulation, Birmingham, December. Offer (1998a). Review of Electricity Trading Arrangements: Background Paper 1 – Electricity Trading Arrangements in England and Wales. Office of Electricity Regulation, Birmingham, February. Offer (1998b). Review of Electricity Trading Arrangements: Background Paper 2 – Electricity Trading Arrangements in other Countries. Office of Electricity Regulation, Birmingham, February. Offer (1998c). Review of Electricity Trading Arrangements: Working Paper on Trading Inside and Outside the Pool. Office of Electricity Regulation, Birmingham, March. Offer (1998d). Review of Electricity Trading Arrangements – Interim Conclusions. Office of Electricity Regulation, Birmingham, June. Offer (1998e). Review of Electricity Trading Arrangements. Office of Electricity Regulation, Birmingham, July. Offer (1999). The New Electricity Trading Arrangements, Vol. 1. Office of Electricity Regulation, Birmingham, July. Ofgem (2000). A Review of the Development of Competition in the Industrial and Commercial Electricity Market. Office of Gas and Electricity Markets, London. Ofgem (2002). Annual Report. Office of Gas and Electricity Markets, London. Ofgem (2003). Review of Competition in the Non-domestic Gas and Electricity Supply Sectors. Office of Gas and Electricity Markets, London. Ofgem (2004a). Security of Supply April to September 2003, Six Month Retrospective Report. Office of Gas and Electricity Markets, London, February. Ofgem (2004b). Domestic Competitive Market Review 2004: A Review Document. Office of Gas and Electricity Markets, London, April. Parkinson, C. (1992). Right at the Centre. Wiedenfield and Nicolson, London. Patrick, R.H. and Wolak, F.A. (1997). Estimating the customer-level demand for electricity under realtime pricing. Paper Presented to the POWER conference, March, Berkeley, University of California Energy Institute. Pollitt, M.G. (1999). The Restructuring and Privatisation of the Electricity Supply Industry in Scotland, mimeo, Cambridge, available at http://www.econ.cam.ac.uk/electricity/people/pollitt/esiscotland.pdf Robinson, G. (2001). The Unconventional Minister: My Life Inside New Labour. Penguin, London. Salies, E. and Waddams Price, C. (2004). Charges, costs, and market power: the deregulated UK electricity retail market. The Energy Journal, 25(3), 19–37. Shuttleworth, G. (1999). The Electricity Industry 1997–1998 (Chapter 2). In P. Vass (ed.), Regulatory Review 1998/1999. CRI, University of Bath, pp. 19–34. Steiner, P. (1957). Peak loads and efficient pricing. Quarterly Journal of Economics, November. Sweeting, A. (2001). Market Outcomes and Generator Behaviour in the England and Wales Wholesale Electricity Market 1995–2000, mimeo, MIT. Turvey, R. (1968). Optimal Pricing and Investment in Electricity Supply. MIT Press, Cambridge. Vardi, J., Zahavi, J. and Avi-Itzhak, B. (1977). Variable load pricing in the face of loss of load probability. Bell Journal of Economics, 8(Spring), 270–288. Vazquez, C., Rivier, M. and Perez-Arriaga, I.J. (2002). A market approach to long-term security of supply. IEEE Transactions on Power Systems, 17(2), 349–357.

This page intentionally left blank

Chapter 5 The Nordic Electricity Market: Robust by Design?1 EIRIK S. AMUNDSEN,1 LARS BERGMAN,2 AND NILS-HENRIK M. VON DER FEHR3 1

Department of Economics, University of Bergen, Bergen, Norway and Institute of Food and Resource Economics, The Royal Veterinary and Agricultural University, Copenhagen, Denmark; 2 Stockholm School of Economics, Stockholm, Sweden; 3Department of Economics, University of Oslo, Oslo, Norway

One of the desirable characteristics of a competitive market is how well it can handle stress, arising from natural or man-made causes. In the case of electricity markets, natural causes, including abnormal weather patterns and/or droughts, particularly in systems, which are predominantly hydroelectric. The Nordic market experienced and survived a severe hydro shortage during 2002–2003. This chapter provides an overview of the Nordic market, focusing on its resilience to handle natural calamities, in contrast to markets such as California, which did not behave well during a similar stressful situation in 2000–2001. Main conclusions are that fears regarding supply security and adequacy are likely to be unfounded. Nevertheless, as inherited over-capacity is eroded, and new market-based environmental regulation takes effect, tighter market conditions are to be expected. It is then crucial that retail markets are fully developed so as to allow consumers to adequately protect themselves from occurrences of price spikes. 5.1. Introduction The Nordic electricity market – encompassing Denmark, Finland, Norway and Sweden2 – is well established by now. Starting in Norway in 1991, regulatory reform gradually spread to Sweden (1996), Finland (1997) and Denmark (2002). In all countries, separation of competitive and monopolistic activities, establishment of independent Transmission System 1

This chapter is an updated version of an article titled: The Nordic market: signs of stress, published in 2005 in the Special Issue of The Energy Journal on European Electricity Liberalization (von der Fehr et al., 2005). The authors are indebted to the International Association for Energy Economics for allowing reuse of this material in the current chapter. The contributions of Perry Sioshansi and Erling Mork, formerly with Nord Pool, are gratefully acknowledged. Thanks are due to Elforsk AB, Market Design and to Renergi, The Research Council of Norway for financial support of work resulting in parts of the present chapter as well as in the aforementioned article. 2 The fifth Nordic country–Iceland–is not interconnected with the others. The term “Scandinavia” is inappropriate, as it only encompasses Denmark, Norway and Sweden.

145

146

Electricity Market Reform

Table 5.1. Background data of the Nordic electricity market. Denmark Installed capacity (MW) Total Hydropower Nuclear power Thermal power Wind power

12,830 11 0 9704 3115

Generation mix Hydropower (%) Nuclear power (%) Thermal power (%) Wind power (%) Consumption and population Total consumption (TWh) Total generation (TWh) Population (million)

Finland

Norway

Sweden

Sum

16,893 2978 2640 11,225 50

28,081 27,676 0 305 100

33,361 16,143 9441 7378 399

91,165 46,808 12,081 28,612 3664

0 0 76 24

18 16 66 0

99 0 1 0

48 28 22 1

51 13 31 4

35.2 43.8 5.387

84.9 79.9 5.220

115.0 107.1 4.565

145.5 132.5 8.976

380.6 363.3 24.148

Maximum system load 2003 (MWh/h) Denmark – West 3780 Denmark – East 2665 Finland 14,040 Norway 19,984 Sweden 26,400 Source: Nordel 2003 Annual Report.

Operators (TSOs) and allowing consumers to choose their supplier have been integral parts of reform. With a long tradition of Nordic co-operation, and with development of the jointly owned power exchange, Nord Pool, the Nordic market is now de facto fully integrated, at least at the wholesale level.3 The Nordic market consists of over 91,000 MW of installed capacity, roughly half of which is hydroelectric, notably in Norway. Statistical details are provided in Table 5.1. Figure 5.1 shows the installed capacity in various countries. By way of background, the Nord Pool’s physical market is a day-ahead spot market where both producers and consumers/distributors submit bids for purchase and sale of electricity for each hour of the coming 24-hour period. These bids are then matched and separate hourly prices are calculated for the next day (midnight to midnight). Capacity constraints are reported by the four TSOs in the Nordic region: Energinet.dk in Denmark (merged from Eltra in West Denmark and Elkraft in East Denmark), Fingrid in Finland, Svenska Kraftnätt in Sweden and Statnett in Norway. Firstly a “system price” is calculated disregarding capacity constraints. This price is used as a reference for cash settlement in the financial derivative market. Then transmission capacities are taken into account. If there are no bottlenecks, then the price for all areas is the same as the system price. However, if there is a bottleneck, then some or all areas may have different prices. Norway is divided into two or more areas (set dynamically by Statnett, the Norwegian TSO), while Sweden, Finland, West and East Denmark are fixed areas. (Note that East and West Denmark have no direct 3

For a detailed map of the Nordic countries showing location and types of plants as a well as a layout of the transmission network, see http://www.nordel.org/Content/Default.asp?PageID⫽125& LanguageCode⫽EN

147

The Nordic electricity market: robust by design? 100

Denmark Finland Sweden Norway

90 80 70

GW

60 50 40 30 20

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

1972

0

1970

10

Fig. 5.1. Installed capacity in the Nordic region. Source: Nordel.

physical connection between them.) In this way, players do not pay explicitly for capacity but bids in the market are used to define the need for various area prices. In 2003 about 25% of hours had a common price across all areas. Often there will be small differences between area prices, and only occasionally will the differences be large. Note also that as the demand side participates there is no demand forecasting. One of the important features of the Nordic market is the evolution of a robust power exchange. As shown in Figure 5.2, Nord Pool’s spot market saw over 40% of Nordic consumption traded in 2004. The remaining physical trade occurs through bilateral agreements or other arrangements. There is also a thriving derivatives market, where futures, forwards and options are traded actively by over 300 participants from both inside and outside the Nordic region. In 2004, nearly 1800 TWh were traded both on and off the exchange. Figure 5.2 shows trading volumes for Nord Pool’s physical market (Elspot, the day-ahead spot market and Elbas, an adjustment market), derivative volumes traded over the exchange’s Financial Market, as well as that which was traded “over the counter” (OTC) or bilaterally but cleared at Nord Pool. Volumes peaked in 2001 and 2002, falling somewhat in the following years due to a severe drought yielding high prices, and the withdrawal of Enron and other major international players from the market. Financial trading volumes have been between 4 and 8 times physical production in the Nordic region. In summary, the Nordic electricity market now constitutes a well-functioning market with the following key design features: ● ● ●

Competition in generation and retailing. Regulation of transmission and distribution. National TSOs responsible for system operation and for running real-time markets (differing from country to country).

148

Electricity Market Reform 3500 3000

TWh

2500

Cleared OTC volumes Financial market traded and cleared at Nord Pool Physical market: Day-ahead spot and adjustment markets

2000 1500 1000 500 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 5.2. Nord Pool market volumes. Source: Nord Pool ASA, Nord Pool Spot AS.

● ● ●

Point-of-connection transmission tariffs. Zonal pricing system of wholesale electricity. Free choice of supplier.

Table 5.2 lists the main events of the Nordic electricity market. In this chapter, a particular focus is put on how the Nordic market was able to withstand a serious shortage of capacity resulting from an unprecedented drought during the 2002–2003 period. This resilience stands in contrast to a number of other markets where less significant perturbations resulted in major mayhem.4 In the second half of 2002, inflow to hydro reservoirs was only 54% of the average of the preceding 20-year period (Bye et al., 2003b). As a result, reservoir fillings were at a record low at the beginning of the low-inflow/high-demand Winter season. Foreseeing tighter market conditions, producers began restricting supply in late Autumn and prices started to rise. The (daily average) spot price peaked at 850 NOK/MWh in January 2003, 2–3 times the normal level. High spot prices feed through to consumers, who in some cases faced increases in electricity bills of 50% or more.5 There was speculation that high prices were the result of abuse of market power, as well as a lack of investment in both generation and transmission in earlier years, and that rationing on a massive scale would be required. As it turned out, no such drastic measures were warranted, as responses from consumers and thermal-power 4

For further discussion of the features of the Nordic market, consult Bergman et al. (1999) as well as Nord Pool. 5 Note that, since many Nordic consumers rely on electricity for most domestic energy needs, including heating, electricity bills tend to make up a considerable share of household budgets. For a typical Norwegian household, annual electricity consumption is around 20 MWh (compared with an average of 3.6 MWh in Britain), while the annual bill would amount to around NOK 14,000 (approximately euro 1700) at a price of 250 NOK/MWh.

The Nordic electricity market: robust by design?

149

Table 5.2. Major milestones in the evolution of the Nordic market. Event

Year

Comments

Energy Act, Norway

1990

Power Exchange, Norway

1993

The Nordic Power Exchange

1994

Regulatory reform, Sweden

1996

Norwegian–Swedish Power Exchange

1996

Regulatory reform, Finland

1997

Further market integration

1998

Further market integration

1999

Regulatory reform, Denmark

1999

Further market integration

2000

Reorganization of Nord Pool

2002

Serious hydro shortage

2002–2003

TGCs

2003

Integration of TSOs

2005

ETS

2005

Nuclear power moratorium in Sweden

2005

Provides legal framework for restructuring and regulation Statnett Market AS established as an independent company The Nordic Power Exchange financial market grows through product development. The first Market Council of the Nordic Power Exchange established Provides legal framework for restructuring and regulation The world’s first multinational exchange for trade in power contracts. The power exchange is renamed Nord Pool ASA Provides legal framework for restructuring and regulation Finland joins the Nordic power exchange market area. EL-EX power exchange in Helsinki to represent Nord Pool Elbas established as a separate balance adjustment market in Finland and Sweden. Western Denmark joins Nord Pool Provides legal framework for restructuring and regulation Eastern Denmark joins Nord Pool as a separate price area Spot-market activities organized in Nord Pool Spot AS owned by all of the Nordic TSOs and by Nord Pool ASA Sharply rising electricity prices with significant demand side response. No intervention from authorities. As the first Nordic country Sweden introduces a TGCs market Energinet.dk in Denmark established by merging from Eltra in West Denmark and Elkraft in East Denmark EU Emissions Trading starts January 1. Norway joins A referendum in Sweden in 1980 resulted in a decision to close down all 11 nuclear reactors in Sweden by 2010. Barsebäck’s second reactor closed down in Summer.

producers balanced the market. Even though prices remained high during most of 2003, market conditions gradually normalized. Some saw the events of 2002–2003 as a warning sign, or indeed as outright proof that the electricity market is flawed. Others consider its performance through this period as evidence that the market has reached maturity and is robust enough to withstand even quite extreme shocks. We tend to lean toward the latter view. Nevertheless, the supply

150

Electricity Market Reform

shock brought to the surface a number of potential weaknesses that warrant careful analysis and which may eventually lead to further improvements in the regulatory framework as well as in other market institutions. After describing the events of 2002–2003 in some detail, we analyze two issues that have attracted considerable attention in the aftermath of these events. The first issue concerns the operation of the retail market; in particular, whether there is sufficient competition on the market, and whether current contractual arrangements are adequate for consumer needs. The second issue concerns generation and transmission capacity; in particular, whether sufficient investment is forthcoming and reasonable levels of supply security and adequacy may be maintained. We conclude this chapter with a discussion of how the implementation of the Kyoto protocol may affect the Nordic electricity market. In particular, we address the effects of the new Emissions Trading Scheme (ETS) of the European Union (EU) and the effects of the emerging Tradable Green Certificates (TGC) system of the Nordic countries. These environmental markets are now making a significant impact on the Nordic electricity market. Hence, in order to understand the functioning of the Nordic electricity market it is essential to recognize the importance of these new environmental markets.

5.2. Coping with a Supply Shock6 The development of the electricity market during the Winter of 2002–2003 was spectacular, with prices reaching unprecedented levels and a constant threat, according to some observers, of rationing on a massive scale. We concentrate our attention on events in the Norwegian segment of the market, where effects were at their most extreme. Figure 5.3 shows the Nord Pool (system) spot price (daily average) and the level of Norwegian hydro stocks7 from the beginning of 2002 to the end of the Summer of 2003. (The average exchange rate for 2002 was 7.5 NOK/euro.) The figure also shows median hydro stocks over the period 1990–2000. From a low level during the Summer of 2002, the spot price rose gradually during the early Autumn. This is normal and reflects the fact that limitedstorage capacity makes it impossible to transfer sufficient water into the high-demand/ low-inflow Winter season to equate prices over the year. However, toward the end of the Autumn the spot price rose steeply and continued to rise well into the Winter, when it peaked at around 850 NOK/kWh. The spot price then fell during the late Winter and Spring, but remained relatively high during most of 2003. The price development reflects the development of hydro stocks. Stocks fell from high levels in the Summer of 2002 to record low levels in the following Winter. The most obvious reason for this unusual development was the extremely dry hydrological conditions with an almost total stop in inflows to reservoir during the normally wet weeks of the late Autumn.8 As shown in Figure 5.4, which compares weekly inflow to Norwegian hydro reservoirs during 2002 with yearly averages over the period 1970–1999, the year 2002 actually started out as rather wet. Until the Summer, inflow was consistently above the historical average; 6

This section draws extensively on Bye et al. (2003a); see also Bye (2003), Bye and Bergh (2003) and Bye et al. (2003b). 7 The Nordic hydro generation capacity is almost entirely located in Norway and Sweden. The Swedish hydro stocks followed a parallel development to the Norwegian. 8 Inflow to reservoirs is at its minimum in the Winter season when most precipitation is in the form of snow, and it reaches its maximum in the late Spring and early Summer when the snow melts. Autumn is usually rather wet, with almost all precipitation in the form of rain, and hence inflow is relatively high in this period also.

151

The Nordic electricity market: robust by design? 900

100 90 80 70 60 50 40 30 20 10 0

800

Percent

600 500 400 300

NOK/MWh

700

200

Spot price

Hydro stocks

24-09-03

13-08-03

02-07-03

21-05-03

09-04-03

26-02-03

15-01-03

04-12-02

23-10-02

11-09-02

31-07-02

19-06-02

08-05-02

27-03-02

13-02-02

02-01-02

100 0

Median stocks

Fig. 5.3. Spot price and Norwegian hydro stocks, actual 2002–2003 and median 1990–2000. Source: Statistics Norway and Norwegian Water and Energy Authorities. 9000

Year 2002 Yearly average, 1970–1999

8000 7000

GWL

6000 5000 4000 3000 2000 1000 0 1

5

9

13

17

21

25

29

33

37

41

45

49

Fig. 5.4. Weekly inflow (GWh) to Norwegian hydro reservoirs. Source: Norwegian Water and Energy Authority.

indeed, in the 24 first weeks of 2002 inflow was 14 TWh, or nearly 20%, above average. However, in early Autumn inflow fell below normal levels and from October onwards it more or less dried up completely; during weeks 38–48 inflow was 9.3 TWh below average. It has been argued that the fall in hydro stocks could have been avoided if generators had restricted supply at an earlier stage. However, with the very high levels of stocks in the early Autumn there was apparently a real risk that – with a wet Autumn – reservoirs would have become so full that water would have been spilled and wasted. Figure 5.5 shows actual total hydro stocks, as well as projected levels as seen from week 30. With maximum inflow reservoirs would have become completely full. Given differing levels of stocks, inflow patterns and constraints on output, the risk that any individual

152

Electricity Market Reform

Percent

100 90 80 70 60 50 40 30 20 10 0

Actual Projected minimum Projected maximum Projected mean 1

5

9

13 17 21 25 29 33 37 41 45 49 Week

Fig. 5.5. Norwegian hydro stocks 2002, actual and projected from week 30. Source: Statistics Norway and Norwegian Water and Energy Authorities.

reservoir would be filled up was likely to be higher than suggested by these figures. As it turned out, such a wet outcome did not occur; instead, inflow and hydro stocks were in fact close to the lowest projected level. This evidence would seem to be consistent with a view that generators acted rationally upon information available at the time. On the other hand, the evidence is also consistent with the view that hydro generators were overly anxious to tap their reservoirs, possibly with the intention of pushing up prices, which, at the time, looked to remain at modest levels during the Winter. In any case, the mere suspicion that such anti-competitive practices were pursued may well have been influential in determining the tough stance that the Norwegian Government took on attempts by the dominant (and state-owned) generator Statkraft to buy up more of its smaller Norwegian rivals. Another possible explanation for the price spike, popular among a number of commentators, was that a lack of investment over a long period of time had eventually lead to undercapacity and a consequent imbalance between supply and demand. It is certainly true that investment levels had been low for a number of years, but whether this was a sign of market imperfection is questionable. We discuss the investment issue in some detail below. Here we only point to the fact that even though prices reached record levels during 2002–2003, they had been well below levels that would make new investment profitable for most of the preceding 10–15 years. Forward contract prices were also low. For instance, in 1999 prices in forward contracts covering the year 2002 were around 150 NOK/MWh, well below the 200 NOK/MWh estimate at the time of unit costs for new gas-fired plants. Forward prices gradually increased in subsequent years, although in 2002 prices covering the year 2006 were still not higher than 180 NOK/MWh. There consequently seems to be little support for the view that generators had not seized on profitable investment opportunities. A criticism along similar lines concerned investment in transmission capacity.9 Some commentators argued that lack of investment in transmission capacity, both within and between 9

A completely opposite, but nevertheless quite popular view was that high prices in Norway were caused by excessive interconnection with neighboring countries, leading to a drain on hydro resources and consequent shortage prices. While there was certainly some truth in this view, it seemed to overlook the fact that in the face of a severe, negative supply shock prices would reach even higher levels without access to imports. We do not discuss the possible (protectionist) implications of this view.

153

The Nordic electricity market: robust by design? 700 Households Service industries Manufacturing industries

600 NOK/MWh

500 400 300 200 100

2004:1

2003:4

2003:3

2003:2

2003:1

2002:4

2002:3

2002:2

2002:1

2001:4

2001:3

2001:2

2001:1

2000:4

2000:3

2000:2

2000:1

1999:4

1999:3

1999:2

1999:1

0

Fig. 5.6. End-user prices (excluding taxes and network tariffs), quarterly observations, 1999–2004. Source: Statistics Norway.

the Nordic countries, had allowed the development of areas where severe shortages were bound to arise. Again, it is true that there had been relatively little investment in transmission capacity over a number of years, and that the 2002 supply shock leads to long periods of time in which the market was effectively segmented. In particular, import capacity to Norway, and export capacities from the thermal-dominated systems in Denmark and Finland were constrained during much of the Winter of 2002–2003. Nevertheless, it is not clear that this was a sign of deficiencies in the transmission network. On the one hand, the Nordic countries are highly interconnected, and most of the time the market is fully integrated at a single market-wide price. On the other hand, bottlenecks are becoming more frequent, and, in combination with increasing levels of market concentration, there is a worry that the result may be imperfect competition and inefficient market outcomes. We return to this latter issue below. We also return to the issue of transmission investment in connection with our discussion of supply security and adequacy. The high spot prices feed through to end-user prices. Figure 5.6 shows quarterly observations of the energy element in average retail prices from early 1999 to early 2004 for households, services industries and manufacturing industries, respectively. Retail prices shot up at the end of 2002 and soon reached unprecedented levels. Prices paid by households tended to increase more than those paid by industrial consumers. The difference seems to be explained by the different composition of contracts in the various segments of the market. Most household consumers have the so-called “variable-price contracts”, according to which retailers can change the price with a few weeks notice. As of the first quarter of 2003, 85% of household consumers had such contracts; another 7% were on spot-price contracts (with the retail price directly linked to the Nord Pool spot price) and only 8% on fixed-price contracts (see Fig. 5.7). This is different from industrial consumers, especially in the manufacturing industries, who tend to rely more on long-term, fixed-price contracts. In the first quarter of 2003, 55% of consumers in the manufacturing industries and 22% of consumers in service industries had fixed-price contracts. The corresponding figures for spot-price contracts were 35% and 53%, and for variable-price contracts 10% and 24%. Consequently, industrial consumers were less exposed to price increases than were households. There was a general move from variable-price contracts to fixed-price contracts in the wake of the events of the Winter of 2003. This trend seems now to be reversed (maybe because the

154 100 90 80 70 60 50 40 30 20 10 0

2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1 2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1 2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1

Percent

Electricity Market Reform

Households

Service industries

Variable price

Spot price

Manufacturing industries Fixed price

Fig. 5.7. Contract shares, quarterly observations. Source: Statistics Norway.

memory of the price spike is starting to fade?). Interestingly, there is little interest among household consumers, as opposed to industrial consumers, for the so-called “spot-price” contracts, in which the retail price is linked to (an average of) the Nord Pool Elspot price. There is ample evidence that spot-price contracts perform consistently better than variable-price contracts in the longer term; in particular, it would seem that competition between suppliers of variable-price contracts is not always entirely effective. The reason households have not embraced spot-price contracts may be (an erroneous) view that these contracts, being linked to the highly variable spot price, are somehow more risky than variable-price contracts. Increases in end-user prices had a considerable impact on demand. Roughly speaking, demand may be seen as consisting of three segments: the very flexible boiler segment (approximately 5% of the total), the heavily contracted power-intensive industry (approximately 30%) and the rest (approximately 65%). Demand from the boiler segment, which can easily switch between oil and electricity, fell sharply when prices started to rise in October 2002 and remained low during the Winter; all in all, electricity consumption by boilers over the period November 2002 to May 2003 was around one-third of that of the corresponding period in 2001–2002. In the energy-intensive industries, some plants stopped production, but the overall response was relatively small, probably less than 5% (Bye et al., 2003a).10 In the remaining segment, households and other industry, temperature-adjusted demand fell by 7% over the November–May period compared to the year before; given an average increase in end-user prices of 30%, this corresponds to a price elasticity of 0.23. The experience in Norway may be contrasted with that of the other Nordic countries. Although wholesale prices moved more or less in parallel, retail prices were much less affected in these countries. This would seem to be explained by the fact that retail markets differ, particularly in the availability and composition of contracts, but also in market structure and the extent of competition. In Denmark and Finland, where fixed-price contracts 10

The response seems small in comparison to estimated reservation prices for the power-intensive industry. The industry has the option of moth-balling plant and selling power, obtained of preferential terms determined by the Norwegian Parliament, in the spot market (Bye and Larsson, 2003). The lack of response may be due to uncertainty about price developments, long stop and start-up times and the risk of undermining popular and political support for the industry.

The Nordic electricity market: robust by design?

155

dominate, domestic consumers were much less exposed to price increases than in Norway. In Sweden, there is a greater variety of contract types, although the incidence of long-term, fixed-price contracts is higher than in Norway. Moreover, as we discuss below, there seems to be less competition among Swedish than among Norwegian retailers. Also, in Sweden retail prices reacted much less than in Norway. As a result, the demand response was much less in these countries than in Norway. 5.3. Market Integration and Retail Competition The events of 2002–2003 cast light on a number of potential problems, including concentration and scope for market power, contractual coverage and exposure of consumers to price risks, investment and its impact on supply security, as well as bottlenecks in the transmission network and segmentation of the market. Below we focus on the retail market (this section) and issues concerning supply security and adequacy (next section). We also discuss how the introduction of new, market-based instruments for regulation of environmental emissions may impact on the performance of the electricity industry (penultimate section). The establishment of Nord Pool and the elimination of border tariffs between the Nordic countries were key elements in a strategy aiming at an integrated Nordic market for electricity. The success of this strategy may be measured by the degree of wholesale and retail price equalization between the different “price areas”.11 Obviously, an uneconomically large transmission capacity would be required if transmission constraints were to be eliminated, enabling wholesale prices to be equalized across all areas at all times. However, significant and persistent deviations between area prices would imply that the Nordic market consists, in effect, of a set of national or regional electricity markets. As we show immediately below, the wholesale market appears to be strongly integrated, with prices in different areas diverging for shorter periods only. However, as mentioned above, retail market prices reacted very differently across the Nordic countries to the 2002–2003 increase in wholesale prices: while they shot up in Norway, the reaction was much more subdued in Sweden, and in Denmark and Finland retail prices hardly changed at all. There are also considerable differences in the level of retail prices, even when one corrects for differences in taxes and network tariffs. Some of these differences can be explained by differences in regulatory regimes. We concentrate our attention on Norway and Sweden, where regulations are similar, but where retail markets nevertheless seem to perform quite differently.

5.4. Wholesale Prices Table 5.3 displays the annual averages of the Elspot system12 and area prices over the period 1996–2003. The figures indicate that deviations between system and area prices have been

11

Whenever interconnector capacity constrains power flows, the Nord Pool market is divided into two or more “price areas”. Sweden is always treated as a single price area, and the same applies to Finland. This is because congestion in the national transmission systems is managed by means of the so-called counter-trade in these countries. In Denmark, the eastern and western parts of the country are physically separated and hence there are always two price areas – East and West. In Norway, segmentation of the market is part of the handling of transmission constraints and the country may be divided into two to five price areas, depending on the demand–supply configuration. 12 The “system price” is calculated under the assumption that there are no transmission constraints. Actual trade is carried out at the system price only when transmission constraints are not binding.

156

Electricity Market Reform

Table 5.3. Elspot system and area prices 1996–2003, annual averages, NOK/MWh.

System Norway, Oslo Norway, Tromsø Sweden Finland Denmark, West Denmark, East

1996

1997

1998

1999

2000

2001

2002

2003

253.6 256.7 251.2 250.6 – – –

135.0 137.5 133.0 135.0 – – –

116.4 115.7 116.2 114.3 116.3 – –

112.1 109.2 119.5 113.1 113.7 113.7 –

103.4 97.7 100.7 115.5 120.7 120.6 –

186.5 186.0 188.6 184.2 184.0 191.2 189.7

201.0 198.5 200.2 206.3 203.8 190.7 213.7

297.5 301.7 295.7 292.8 277.9 268.3 291.7

Source: Nord Pool. Table 5.4. Number of hours with complete Elspot price equalization 1997–2002.13

Number of hours Share of time (%)

1997

1998

1999

2000

2001

2002

5201 59.4

3825 43.7

3788 43.2

1703 19.4

4487 51.2

3076 35.1

Source: www.seef.nu.

quite small. Except for the years 2000 and 2003 – when the supply of hydropower, especially in Norway, significantly deviated from normal levels – the Nordic electricity market appears to be reasonably close to being a “single market”. However, small discrepancies between annual averages may hide short-term variations in different directions and is thus only a very crude indicator of the degree of price equalization. As can be seen from Table 5.4, the number of hours in which the entire market has been integrated (i.e. when the system price has been exactly equal to all area prices) is below 60%. The corresponding figures for Finland–Sweden are in the range 75–100%, for Norway (Oslo)–Sweden 70–85% (except in 2000) and for Finland–Norway (Oslo)–Sweden 60–85% (except in 2000). Moreover, the share of the time when Sweden has been a “price island” is in the range 0–5%. Thus, in terms of wholesale price equalization the Nordic electricity market would seem to be reasonably well integrated. 5.5. Retail Competition There is complete market opening (i.e. full retail competition) in all the Nordic countries. In some of the countries, such as Sweden, a household consumer may even buy electricity from suppliers in any Nordic country. Given this, the pre-tax retail prices should not differ very much between the four Nordic countries. However, there are obstacles to transactions between suppliers in one country and households and other small customers in other countries. For instance, in order to be able to supply electricity to a Swedish customer located in Stockholm a non-Swedish supplier needs to buy electricity in the Stockholm price area. Moreover, as a buyer in the Stockholm price area the supplier needs to have a contract with a so-called “balance responsible party” (as well as in the home country).

13

Presentation by Dr. Niklas Strand, Swedish Competition Authority, at the Swedish Association for Energy Economics.

The Nordic electricity market: robust by design?

157

As a result of such obstacles, only a sub-set of all retailers in the Nordic countries is actually competing on the Swedish market. In addition, only a sub-set of all Swedish retailers competes outside the geographical area in which they are located. Corresponding situations prevail in the other Nordic countries. Consequently, retail electricity prices need not necessarily be equalized across national borders. In order to shed some light on this issue, we have compared retail prices in Norway and Sweden, the two countries that have operated a common wholesale market since 1996. Retail prices differ between households for several reasons. One reason is related to nonlinearity of price schedules and annual consumption patterns of households. Thus prices paid by households living in a single-family houses with electric heating, typically consuming around 20 MWh/year, are lower than the prices paid by households using electricity only for lighting and electrical appliances, consuming as little as 2 MWh/year.14 Another reason for retail price differences is that customers can choose between fixed-price contracts (with different duration) and variable-price contracts, and that the prices charged for these contracts may deviate in the course of year. In both Norway and Sweden, the “default contract” – the type of contract that applies for customers who have not actively chosen to change supplier or signed a new contract with the “old” supplier – is a variable-price contract; that is, a contract that allows the supplier to adjust the price (after notifying the customer) and so, in effect, pass on cost increases to customers. A third reason for retail price differences is that individual retailers may adopt different market strategies and offer different combinations of prices and services. In Tables 5.5 and 5.6, annual averages household prices over the period 1997–2003 are displayed for, respectively, Norway and Sweden. The numbers reflect averages of prices offered by all retailers to households with an annual consumption equal to 20 MWh. It is immediately clear that Norwegian and Swedish retail prices differ significantly. No “law of one price” is visible, and the retail prices have been significantly higher in Sweden during most of this period. Thus, with regard to the household market, there seems to be two national rather than one integrated Norwegian–Swedish electricity market. Annual variations of the Norwegian retail prices are quite significant, but also relatively well correlated with variations in Elspot prices (compare Tables 5.3 and 5.5). Thus retail prices fell during the “wet” period 1997–2000, increased in 2001 when precipitation was “normal”, and skyrocketed in 2003 when Nord Pool prices were extremely high. In the case of Sweden, however, the correlation between Elspot and retail prices was rather weak during the period 1996–1999. Instead, average retail prices remained high in 1998 and fell only marginally in 1999, while a quite significant reduction took place in 2000 and 2001. The most obvious explanation for these differences between the development of retail prices in Norway and Sweden are related to switching15 costs at the household level. In Norway, a system of profiling was adopted from the outset and customers could switch to another supplier at no cost; more specifically, there were no requirement to install interval or real-time meters, nor any charges levied. The option of changing supplier at no charge has led to a sizeable swing of customers away from the “old” local suppliers to

14

To some extent this pattern is surprising. On the one hand, it is likely that economies of scale may motivate some quantity discounts. On the other hand, it is obvious that households with electric heating consume electricity primarily during the Winter period when peak generation capacity is used and spot prices are high. As spot prices are sometimes very high during the Winter season, one would expect that the cost of offering the customer insurance against Winter “price spikes”, which is what the retailer does, would be rather high.

158

Electricity Market Reform

Table 5.5. Retail prices, net of taxes, for 20-MWh household consumer according to contract type in Norway 1997–2003, NOK/MWh.

Normal Spot 1 year fixed Average

1997

1998

1999

2000

2001

2002

2003

n.a. n.a. n.a. 210

160 n.a. n.a. 162

151 131 152 152

141 123 144 141

210 193 189 206

205 193 195 203

454 323 287 414

Source: Statistics Norway.

Table 5.6. Retail prices, net of taxes, for 20 MWh household consumer according to contracts in Sweden 1997–2003, NOK/MWh.

Normal 1 year fixed 2 year fixed 3 year fixed

1997

1998

1999

2000

2001

2002

2003

259

251

244

218 178 177 182

225 181 184 186

296 256 253 252

447 397 351 324

Source: Statistics Sweden.

“new” suppliers that offer electricity at lower prices. In other words, there were no significant switching costs protecting the “old” suppliers from competition. In Sweden, on the other hand, costly real-time metering and reporting were required for consumers wanting to change supplier. These regulations were in effect until November 1999, and as a result few households changed supplier or renegotiated their contract with the “old” supplier. After November 1999, a system of profiling has been in place. Consequently, there was a significant reduction of switching costs in Sweden at the end of 1999. The figures in Table 5.6 suggest that retail competition was rather modest as long as switching costs were high. At the same time, the figures indicate that the reduction of switching costs opened up a significant competitive pressure on retail prices. Nevertheless, even with a regulatory regime more conducive to competition, retail prices continue to be higher in Sweden than in Norway, in spite of the fact that retailers procure electricity at the same well-integrated wholesale market. A natural hypothesis is that this is a result of market power.

5.6. Market Structure Traditionally, the major generating companies in Sweden have had rather small shares of the retail market. Thus, although Vattenfall for a very long time has been the single biggest

15

It should be noted that the very high per capita electricity consumption for heating, lighting and household utensils of Norwegian and Swedish households make it profitable to switch electricity supplier when prices differ. This may, however, not be the case in other European countries (e.g. the UK) where electricity is mainly used for lighting and household utensils.

The Nordic electricity market: robust by design?

159

retailing company, its share of the retail market was only around 15% in the middle of the 1990s. However, in the last few years the major generating companies – Vattenfall, Sydkraft and Fortum (formerly Birka) – have bought majority or minority shares in a number of small- and medium-sized retailing companies. In most cases, sellers have been towns and municipalities. Moreover, some of the independent retailing companies that entered the market in 1999, such as the Norwegian oil and gas company Statoil, have since left the market. As a result of these developments the number of retailing companies has been reduced, and the “big three” have become dominant players on the retail market. For instance, if we include retailing companies in which generators own minority shares, Vattenfall is currently serving around 30% of all Swedish customers, while the corresponding number for the “big three” is around 70%. Similar numbers apply to shares of volumes of electricity delivered to final consumers. In Norway, power generation has traditionally been much less concentrated than in Sweden. Except for the state-owned company Statkraft, accounting for some 30% of total Norwegian power generation, generators are small, with market shares of 5–6% or lower. As Statkraft and the second largest company, Norsk Hydro, almost exclusively serve industry and other businesses on long-term contracts, the retail and household market in Norway has been much less concentrated than in Sweden. In recent years, however, significant changes in the Norwegian power sector have taken place. In particular, many companies in local-government ownership have been turned into limited-liability companies, often as part of a process leading up to the sale of ownership interests. Also, larger regional power companies have been established, partly by acquisition and partly through mergers. Furthermore, Statkraft has acquired stakes in several Norwegian power companies. Foreign companies have also acquired some ownership interests in Norwegian companies (notably in grid management and operations and in power retailing). In spite of these developments, however, the Norwegian market, both at the wholesale and retail level, remains less concentrated than its Swedish counterpart. 5.7. Market Power and Price Discrimination So far, the retailing segment of the Swedish electricity supply industry has not been very profitable, and several entrants to the market have had to leave after having suffered significant losses. On the whole, it seems that the costs of the retailing business have been severely underestimated; in particular, it seems that the exposure to price and quantity risks have been more costly than expected.16 But in 2002 the established retailers were able to implement an increase in trade margins without attracting new entrants to the market. The question then is why “the big three” have grown while several independent retailers have left the market. A possible explanation may be that integrated generation-retailing companies are more efficient than independent retailers. In Sweden, legal separation between retailing and distribution is required. This provision has paved the way for significant integration 16

Retailers can hedge Elspot system price risks at Eltermin, and the liquidity of these instruments (futures, forwards and options) traded at Eltermin is high. However, retailers in Sweden have to buy electricity at the relevant Swedish area price, and opportunities for hedging idiosyncratic area price risks are not very well developed. Moreover, retailers can only hedge price risks for a fixed number of MW per hour, while their customers do not have to commit to a certain quantity. Thus retailers are exposed to the risk of having to buy “extra” electricity at spot-market price during hours when their customers have an unexpectedly high level of consumption.

160

Electricity Market Reform

between generation and retailing. It seems that the integrated companies have a competitive advantage in relation to independent retailers stemming from the lack of efficient markets for hedging against area price and quantity risks. Thus, while all retailers may suffer from a combination of unexpectedly high area prices and consumption levels, the extra costs in the retailing business become extra revenues in the generation business for the integrated generation-retailing companies. If these hypotheses could be verified they clearly point at an unexpected effect of the legal separation requirement in Sweden. The intention was to stimulate retail competition by preventing cross-subsidization between distribution and retailing. This objective may have been attained, but recent developments suggest that allowing vertical integration between generation and retailing also made it possible to exercise market power in the retail market. There also seems to be an element of price discrimination in Swedish retail prices. As seen in Table 5.6, there is a considerable spread between prices in the so-called “normal contract” and prices in fixed-prices contracts or between prices paid by customers with default contracts and customers who have actively switched to a new contract; in particular, prices in default contracts are higher than those in fixed-price contracts. This suggests that Swedish retailers are able to price-discriminate against the default-contract customers. After all, by refraining from actively choosing another type of contract these customers have demonstrated that they are not very price-sensitive and it must be tempting for suppliers to utilize this information. 5.8. Supply Security and Adequacy A notable consequence of regulatory reform in the Nordic countries has been the almost complete halt to investment in generation and, to a lesser extent, in transmission and distribution. At the outset this should be considered a positive feature as this has implied a reduction of an inefficient over-capacity in generation inherited from the past. Figure 5.8. shows installed generation capacity in the Nordic market since 1980. Over the 10-year period preceding the first regulatory initiative, from 1980 to 1990 (the year before the new Energy Act took effect in Norway), generation capacity grew by 30%. Over the next 10-year period, from 1990 to 2000, installed capacity grew by a mere 3%. Indeed, in 2003 installed capacity was more or less the same as in 1996, the year when regulatory reform was introduced in Sweden: a fall during 1998–1999 was only reversed by subsequent increases in recent years.17 The stagnation in capacity growth cannot be explained by development of demand. Admittedly, demand did not grow at the pace experienced in the early 1980s and before, but gross consumption continued to grow at a more or less constant rate of 1–1.5% per year. As a consequence, over-capacity inherited from the pre-reform era has gradually being reduced, if not entirely eliminated. Comparing generation capacity and maximum system load, we find a similar picture, although the trend is perhaps not as pronounced. Nevertheless, the capacity margin, defined as the excess of installed generation capacity over maximum load, reached it lowest level in 2001. The development of generation capacity must be seen in relation to stricter regulatory policies, arising mostly from environmental concerns. Although a considerable amount of undeveloped hydro capacity still remains, it is unlikely that many more new hydro sites will be developed. Nuclear power, traditionally important in both Finland and Sweden, has long been viewed with great skepticism (although the Finnish Parliament recently 17

This movement seems to be explained largely by plants that were first mothballed and then re-opened.

161

The Nordic electricity market: robust by design? 450

100 90

400

70

350

60

300

50

Installed capacity Gross consumption Maximum system load

40

GWh

GW

80

250 200

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

30

Fig. 5.8. Capacity, consumption and system load. Source: Nordel.

approved the building of a new nuclear power plant). Increasing concerns about air pollution has lead to strict regulations, not only on coal- and oil-fired power plants, but also gasfired plants. Notwithstanding the importance of environmental regulation, it would seem that regulatory reform – with the abolition of monopoly rights, integration of markets and development of market places – has been the most important factor in determining generator investment behavior. Market-based competition not only reduced the prices, but also turned the focus of market participants toward profitability. Indeed, in an industry traditionally committed to a public service ethos, regulatory reform legitimized a more “capitalist attitude”. The greater emphasis on profits lead to company restructuring and mergers, as well as to increased efficiency (e.g. employment in the Norwegian electricity industry fell from almost 20,000 in 1993 to below 13,000 in 2002). Specifically, electricity supply companies increasingly required returns on investment in line with those obtained in other industries. Figure 5.9 shows return on capital in the Norwegian electricity industry since 1990, the year before deregulation.18 Over this period, rate of return has averaged 5.5%, only half that achieved in the Norwegian manufacturing industry. Due to the existence of a “resource rent” in a hydro-dominated electricity industry, we would expect the average rate of return to exceed that on a marginal plant. Consequently, it seems relatively safe to conclude that new investment in generation capacity could not have achieved reasonable levels of profitability over this period. Nevertheless, even though low levels of investment seem to have been a rational response to prevailing market conditions, the question remains whether investment will be forthcoming as the earlier over-capacity is eroded and market conditions become tighter. In other words are there reasons to believe that market imperfections will inhibit investment and undermine the future performance of the industry? To answer this question, it is important to distinguish between two related, but nevertheless entirely different, concepts: supply security or balancing consumption and generation

18

Rate of return is measured according to National Accounts as operating surplus relative to the value of capital employed.

162

Electricity Market Reform 18 Electricity industry

16

Manufacturing industries

14 Percent

12 10 8 6 4 2 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Fig. 5.9. Return on capital, 1990–2002. Source: Statistics Norway.

on a continuous basis within existing capacity limits; and supply adequacy or ensuring optimal capacity investment by balancing willingness to pay for new capacity against its cost. In other words, while the supply-security issue is short term and mainly concerns system operation, the supply-adequacy issue is long term and concerns the evolution of capacity in relation to consumption. In interpreting the balance of consumption and generation one must be aware of transmission constraints that limit the range of generators that can meet demand in a given location. Below, we discuss these issues in some detail.

5.9. Supply Security: Balancing Demand and Supply As is well known, the problem of continuously balancing consumption and available generation arises from specific features of electricity markets, including the need for electrical equilibrium at all times, unexpected variations in demand and supply, limited possibilities for establishing and transmitting adequate price signals to market participants on a continuous basis, and limited short-run response by market participants to price signals. The gain from increasing supply security is associated with a reduction in the costs of rationing. A rationing event occurs when, at prevailing prices, the desired demand and/or supply of market participants cannot be satisfied and hence their decisions have to be constrained. For example, if demand exceeds supply at prevailing prices, either additional supply (if available) has to be ordered onto the system, or the consumption of one or more consumers has to be forced down. Rationing may be voluntary or involuntary. One example of a voluntary rationing agreement is contracts for operational reserves, by which the system operator obtains the right to call on additional supplies when needed. Another example is long-term load-shedding contracts between consumers and their suppliers (or between consumers and the system operator), by which consumers, upon conditions that have been agreed in the contracts, can be called on to reduce their load. Involuntary rationing typically occurs by zonal interruptions of power supplies. Supply security cannot be ensured by capacity investment alone, although a larger capacity may reduce supply-security problems. Optimal utilization of existing generation capacity, whatever its level, involves setting prices that allow for the highest possible degree of

The Nordic electricity market: robust by design?

163

capacity utilization while at the same time securing sufficient reserve margins. Having more capacity available essentially means that prices will remain at lower levels so that more demand is encouraged and a high level of capacity utilization is ensured. Conversely, when capacity is limited prices will increase so as to reduce demand. Provided generation capacity is optimally used, on longer time scales demand for capacity will follow available capacity and system reserves will not be directly linked to total capacity. Obviously, in the Nordic system, with a large incidence of hydropower, there will be sustained periods of time in which the energy balance is tight. In such situations, a continued balance between demand and supply requires that prices rise, as happened during the Winter of 2002–2003. Note that the price rise would have been less in that event if more of demand had been exposed to the actual cost of electricity: the reliance on fixed-price contracts in large segments of the market exacerbated the supply deficit and pushed wholesale prices higher than they otherwise would have been. On the other hand, it is conceivable that one may in the future experience an even more severe reduction in inflows, with even higher prices as a result. Nevertheless, if, by adjusting price levels, demand can be scaled to be (on average) aligned with existing capacity, the supply-security problem is not associated with the level of demand (relative to existing capacity) per se but rather with variations in demand (and capacity availability). Given this, it must be noted that, in a hydro system, even if the energy balance is tight there is generally a large amount of (power) capacity available. Since hydro plants are typically constructed so as to be able to handle peak inflow levels, and adjustment of output is extremely cheap, the ability to deal with short-term variations in the system is not necessarily impaired in dry periods. To sum up, it is not at all clear that the Nordic market is particularly vulnerable to supply insecurity, even if market conditions were to become tighter in the future. Indeed, the technology mix, and the mere size of the market, would seem to allow for a very high degree of supply security. Moreover, regulations are in place, which should provide Nordic system operators with the tools they need to balance the system. Take the Norwegian TSO Statnett as an example. Firstly, rights and responsibilities have been clearly set out in specific regulations; in particular, Statnett is responsible for system balance and has the right to make the necessary contractual arrangements with market participants to achieve this task. Secondly, marketbased institutions have been set up to achieve balancing in a cost-effective manner. Most importantly, Statnett runs a (near-to) real-time balancing markets in which balancing services are sourced on a short-term basis. Furthermore, at regular intervals Statnett procure strategic reserves, partly in the form of contracts for interruptible demand. All in all, these instruments should be sufficient to guarantee that balancing is achieved at reasonable costs. Studies have indicated that the efficiency of system operations would be further enhanced by tighter co-operation between of system operators (Bjørndal and Jørnsten, 2001; see also von der Fehr et al., 2002). Developments in this direction have recently taken place, with integration of the national balancing markets (Nordel, 2003). However, these efforts are probably not sufficient, and further gains may be had from optimizing the system as a whole, including lower overall reserve margins and increased transmission capacity. 5.10. Supply Adequacy: Optimizing Capacity Investment The supply-adequacy problem essentially consists of two elements, namely ensuring an optimal level of overall generation capacity and an optimal mix of different generation technologies.

164

Electricity Market Reform

An optimal level of overall capacity is characterized by equality between willingness to pay for new capacity and the cost of such capacity. In other words, a situation of underinvestment in generation capacity would be characterized by investment not forthcoming even though the willingness to pay for the associated increase in output is more than sufficient to cover its cost. Similarly, a situation of over-investment would be characterized by the cost of marginal capacity units exceeding consumer willingness to pay for the associated output, a situation well known from the history of the Nordic electricity industry. An optimal mix of generation technologies is characterized by the minimization of costs of satisfying a given consumption profile. Cost-efficient operation requires a mix of technologies with different variable to fixed cost ratios. At one extreme, low-variable/high-fixed costs technologies, such as hydro, nuclear and conventional thermal, operate continuously as base-load units; at the other extreme, high-variable/low-fixed costs technologies, such as small gas- or oil-fired units, are used for demand peaks only. An optimal capacity mix balances the gains from reducing variable operating costs by having more base-load units available against the higher-fixed costs of such units. At the moment, there would seem to be two major concerns regarding supply adequacy in the Nordic market: firstly, whether investment incentives are sufficient to allow overall capacity to expand at a reasonable pace and secondly, how the generation park will be affected by the introduction of new environmental regulation initiatives. Here we focus on the first issue of general investment incentives and leave the latter issue for the next section. In the hydro-dominated Nordic market, wholesale prices swing from year to year, depending upon hydrological conditions. However, these fluctuations in prices tend to average out. Indeed, judging from prices in forward contracts, which are traded up to 3 years ahead on Nord Pool, expected prices tend to be quite stable, evolving slowly in reaction to changes in underlying fundamentals.19 As at late 2004, contracts for 2 and 3 years ahead are trading at around 250 NOK/MWh (31 euros/MWh), a level that would make investment in conventional gas-fired plant approximately break even. It would seem that these prices are in fact sufficient to attract investor interest. The Finnish Government and the EU Commission recently approved a new 1.6-GW nuclear power plant, expected to be in operation from 2009. In Norway, Statoil recently unveiled plans to build a 860-MW gas-fired plant and a 280-MW co-generation unit at its industrial plants at Tjeldbergodden and Mongstad (landing points for North-Sea gas). Gas-fired generation has been a source of controversy in Norway, leading to the downfall of at least one government, and it remains to be seen whether the Statoil projects will be approved. However, whatever the outcome, it would seem that generation investment in the Nordic market is not so much a question of commercial, as of political, will.20 Transmission investment has been fairly limited for a number of years. There seems to be two main issues: firstly, how regulation of the individual TSO affects investment incentives and secondly, the ability of national TSOs to solve co-ordination problems associated with investment in interconnector capacity. In the case of the Norwegian TSO, Statnett, there appears to be no lack of will to invest. Given that costs of new investment will in effect be 19

Hence, due to the well-functioning market for forward contracts and the fact that the Nordic countries are integrated within and beyond, seasonal and annual swings of precipitation in a hydrodominated country like Norway (close to 100%) do not result in similar swings of prices and threats of rationing as in countries like Brazil and Colombia (see other chapters in this book). 20 Uncertainty surrounding the future of the Swedish nuclear capacity is another important example of how the political process interacts with investment incentives. Investment in renewables, such as wind power and generation based on burning of biomass, is critically dependent upon government support.

The Nordic electricity market: robust by design?

165

passed on to network users, Statnett has shown considerable willingness to undertake new projects. Were it not for the more restrictive view taken by Norwegian regulatory authorities, more transmission capacity would indeed have been built.21 The regulatory regimes differ somewhat between the Nordic countries, which may explain why other TSOs have been more reluctant to invest, particularly in new interconnector capacity. However, Nordel, the association of Nordic TSOs, has taken several initiatives both to resolve problems associated with transit (Nordel, 2001), as well as producing plans for co-ordinated expansion of the Nordic transmission network (cf. the “Nordic Grid Master Plan” described in Nordel, 2003). These initiatives would seem to go a long way in resolving transmission problems, but, again, it would seem that regulatory and political will, rather than commercial will, is going to be decisive. 5.11. Emissions Trading In the following we address the effects of the new ETS of the EU and the effects of the emerging TGC system of the Nordic countries. These environmental markets are now making a significant impact on the Nordic electricity market. Among the Nordic countries, Denmark, Finland and Sweden will, as EU Member States, have to implement the ETS,22 while Norway, not being a member state, is not obliged to do so. However, Norway has recently redesigned its former national (more ambitious) ETS proposal to better fit the EU Directive, and Norway seeks in this way to co-operate and participate in the EU arrangement on an equal footing with the other Nordic countries. A possible outcome of this process is that Norway adopts the EU Directive and thus commits to the rules and regulations of the ETS. Table 5.7 gives numbers for CO2 emission in electricity and heat supply. The levels of CO2 emissions are very different. Denmark, with its relatively small electricity industry, primarily based on thermal power, has the highest level of emissions, whereas Norway, with a relatively large electricity industry, almost exclusively based on hydropower, has very little emissions. The simple idea of the ETS is that the permit price will function as a cost increment of using a CO2 emitting resource, with the increment being in proportion to how much that resource emits per unit used. As a result, input substitution is expected to take place in electricity generation, away from coal and gas power toward hydro, wind and nuclear power. Hence it is to be expected that, in the Nordic countries, the ETS will result in an increase in Table 5.7. CO2 emissions in public electricity and heat supply in 1998 (million tons). Denmark 27.8

Finland

Norway

Sweden

18.2

0.2

7.9

Source: IEA (2000). 21

Statnett recently failed to get approval for a sub-sea link to the UK. It would seem that the business case for this project was indeed weak (Aune, 2003). Statnett is however going to undertake considerable investments in the Norwegian network over the coming years (see www.Statnett.no), and the Dutch regulator has now cleared a 700 MW cable to Norway (NorNed). 22 The ETS and its effects on generation are described extensively in the chapter on Germany in this book.

166

Electricity Market Reform

the relative importance of electricity based on “clean” power, accompanied by increasing electricity prices and a relative increase in the export of electricity from countries endowed with clean natural power resources, such as Norway. This is, indeed, what earlier studies of the effects of a common Nordic market for emission permits show; see Amundsen et al. (1999), Unger and Alm (2000) and Hauch (2003). It should be noted that these studies show that the introduction of a common Nordic electricity market has in itself lead to considerable reductions in CO2 emissions. The reason for this is that increased levels of electricity trade lead to substitution of low-cost and emission-free hydro power for electricity generated by high-cost emission-intensive sources.23 Nevertheless, introducing a common Nordic market for emission trade will provide an additional gain. The reason is that equalization of both marginal generation costs (resulting from trade in electricity) and marginal abatement costs (resulting from trade in permits) is more efficient than equalization of either of these alone.

5.12. The Nordic Market for Green Certificates In recent years, a Nordic market for TGCs has been under development in order to facilitate the use of the so-called green electricity, that is electricity generated by wind-power plants, new small hydro-power plants and power plants using bio-fuels. Sweden introduced its TGC system on May 1, 2003 and Norway will follow suit later. It is the intention that Norway and Sweden will start trading TGCs. The Nordic TGC system was to a large extent designed in Denmark and plans existed for its introduction already in 2001. Although the legal foundation for implementation was in place in Denmark, the TGC system has been put on hold, and no decision has been made as to when it will be implemented. Finland has no immediate plans of introducing a TGC system but does participate in the European Renewable Energy Certificate System (RECS), as do the other Nordic countries.24 Sellers of TGCs are producers of electricity using renewable sources. They are issued with a number of certificates corresponding to the amount of electricity they feed into the electricity network. Buyers of TGCs are consumers/distribution companies that are required by the government to hold a certain percentage of certificates corresponding to their total consumption/end-use deliveries of electricity. The percentage requirement functions as a check on total electricity consumption, as the total number of certificates available is determined by the total capacity of renewable technologies. The TGCs are thus seen as permits for consuming electricity. The system implies that producers using renewable energy sources receive both the wholesale price and a certificate for each kWh fed into the electricity network. In a

23

This result is, however, sensitive to the relative cost of electricity generation for the various technologies employed and to future market development. For instance, enlarging the power market to include Northern Germany may well result in larger imports of low-cost, carbon-intensive coal power and lead to more CO2 emission. In such a setting, the value of introducing ETS will be significant. The introduction of ETS may lead to an increase of end-user prices and reduced demand for electricity. Hence, even though ETS will stimulate electricity exports from countries rich in clean power resources, reduced demand for electricity may still dampen electricity transmission between countries. 24 The European RECS facilitates many support schemes for green energy, rather than being a support scheme itself. It is not restricted by national boundaries. RECS provides a mechanism for presenting production of a megawatt-hour of renewable energy by a unique certificate which can be transferred from owner to owner before being used as proof of generation, or exchanged for financial support (http://www.treckin.com)

The Nordic electricity market: robust by design?

167

competitive equilibrium, therefore, marginal generation cost of green electricity is equal to the marginal generation cost of non-green electricity plus the subsidy element (i.e. the TGC price). Hence, for the producer, the marginal cost of providing a kWh of green electricity is the same as that of non-green electricity, wherefore both technologies will be involved in power generation at the margin. In this way, the TGC system is supposed to stimulate new investments in green electricity generation. Analytically, the TGC system implies that the end-user price of electricity becomes equal to the wholesale price plus a percentage of the certificate price (assuming zero distribution costs), determined by the holding requirement. In a competitive electricity market, end-user price is also equal to a linear combination of the marginal generation cost of green electricity and the marginal cost of non-green electricity, with the percentage requirement as the combination weight. Hence a change of the percentage requirement will affect the relevant marginal generation cost. However, as the system is founded on a percentage basis, it is not necessarily true that an increase of the percentage requirement leads to more green generation capacity, though it will lead to less non-green generation capacity (Amundsen and Mortensen, 2001, 2002).25 5.13. Joint Effects and Compatibility of ETS and TGC Both the ETS system and the TGC system will affect CO2 emissions from electricity and heat generation. In this sense, these are two broad market-based measures (on top of other measures like emission standards, direct subsidies for renewables, etc.) to obtain the goal of reducing CO2 emissions (other considerations, such as supply security and infant industry protection also motivated the introduction of the TGC system). In order to provide an indication of the possible effects of these measures, we refer to some results from a simulation study of the Nordic electricity market focusing on Sweden, which, as mentioned, has already introduced a TGC market (Bergman and Radetzki, 2003). The introduction of the TGC system in Sweden (based on a 7% requirement) will lead to certificate prices at the stipulated upper price bound and an increase in the production of green electricity by 10 TWh. However, net export to the other Nordic countries will increase by 5.2 TWh. This somewhat surprising result is explained by the high price of certificates and the resulting low net cost of generating green electricity in Sweden. Hence the introduction of a TGC system in Sweden significantly affects investment decisions in the electricity industry. However, since in these simulations Sweden is assumed to be the only country applying a TGC system, the effect on the common Nordic electricity wholesale price is rather small. With a joint Nordic ETS system in place (or, equivalently, a joint CO2 tax), Bergman and Radetzki calculate that electricity consumption in Sweden would be reduced by an amount corresponding to the growth in demand over the period 2001–2010 which would otherwise have occurred (i.e. without such a scheme). Hence, CO2 emissions from the electricity sector are reduced at the expense of electricity consumption. However, as the cost increase is even higher in Denmark and Finland, the competitive position of the Swedish electricity industry improves and Sweden begins to export electricity to Denmark and Finland.26 25

Potentially, the TGC system involves some serious problems. For instance, certificate prices may be extremely volatile if wind power constitutes a large part of the renewable technologies. Also, problems of market power (i.e. gaming on the electricity and Green Certificate markets) may be severe and lead to a collapse of the system (Amundsen and Nese, 2004). 26 Broadly speaking, these conclusions seem to be in line with other model simulation studies of the joint effects of ETS and TGC systems in the Nordic countries; see Hindsberger et al. (2003) and Unger and Ahlgren (2003).

168

Electricity Market Reform

As for Sweden, the introduction of the TGC system in Norway is targeted at stimulating power generation from wind and bio-fuel sources. Norway is already well endowed with environmentally friendly hydro-power resources, but the construction of new large hydropower plants has more or less come to a halt, due to the lack of sites for large-scale developments. This is in part due to politically imposed environmental constraints and conflicting interests of land use. The emerging alternative is large-scale gas-power plants. The introduction of the ETS system will reduce the profitability of planned gas-power plants (unless they are not exempted from the ETS system). However, with the present ambition of the ETS system it is not quite obvious that the resulting increase in the wholesale price will be sufficient to stimulate the required investment in electricity plants based on wind power and bio-fuel. Thus, in order to realize these plans a sizable stimulus, such as a TGC market (or plain subsidies for that matter), is called for. Although the two measures work toward the same end (i.e. of reducing CO2 emissions) they are not quite compatible. For instance, under the ETS system, stricter emission constraints will lead to increasing permit prices, increasing generation costs for non-green power and thus increasing wholesale prices of electricity. Remuneration to generators of green electricity (i.e. the wholesale price of electricity plus the value of a TGC) will however decrease. The reason for this somewhat paradoxical result is the particular construction of the TGC system, whereby an increase of the wholesale price by 1 cent results in a reduction of the TGC price by several cents (depending on the size of the percentage requirement). The increase in the wholesale price, following a higher permit price, leads to a corresponding reduction of the margin between the end-user price and the wholesale price. This margin is equal to the TGC price multiplied by the percentage requirement. Hence if this margin is reduced by 1 cent, and the percentage requirement is 20%, the TGC price will be reduced 1/0.2 ⫽ 5 cents. The remuneration to a green producer (the sum of the wholesale price and the TGC price) is therefore reduced. The equilibrium effect of stricter emissions constraints is a reduction of both nongreen and green electricity generation, such that the percentage requirement is still satisfied (for further explanation, see Amundsen and Mortensen, 2001). Investigating this problem in a numerical model, comprising both an ETS system and TGC system for the Nordic countries, Unger and Ahlgren (2003) identify a negative effect of stricter CO2 constraints on green electricity generation capacity. They conclude that this effect is probably not very large in the longer run, as the effect of constraining CO2 emissions further will have only a small effect on electricity wholesale price. 5.14. Conclusion Since the deregulation process started in Norway in 1990, the Nordic electricity market has expanded steadily both in terms of countries and regions included and in terms of contracts traded. For the first many years the market developed rather smoothly without any significant calamities to be dealt with and the market, thus, started to resemble as success. This impression was reinforced after the critical dry Winter season of 2002–2003 when consumers, used to a sustained period of low prices, became confronted with sudden and sharp price increases. In general it seems fair to say that the market withstood this test of robustness and handled the supply shock rather well: prices adjusted rapidly and both demand and supply responded. Drastic measures such as rationing, foreseen by some and feared by many, were never warranted. As such, it would seem that the market has reached maturity. Nevertheless, this event also made it clear that potential problems still exist. As the overcapacity of the pre-reform era is vanishing, the market is becoming increasingly tight. One must consequently be prepared, not only for higher, but also for more variable prices.

The Nordic electricity market: robust by design?

169

Tighter market conditions also means that system operation will become more challenging, particularly ensuring the availability of sufficient reserves of generation capacity and, most importantly, interruptible demand so as to achieve system balance at all times. Nordic system operators seem on the whole to be well prepared, although further co-operation (integration) of system operations could potentially enhance efficiency. Tighter market conditions and increasing prices signal a need for capacity additions. So far, there is little evidence that generators do not react to such signals; indeed, at the time of writing capacity additions are either being planned or are already under way. However, uncertainty surrounding the political will to accept growth in electricity generation and consumption – exemplified with the lack of clarity about new measures for regulating environmental pollution – may undermine the willingness to invest. From a supply-adequacy point of view, it is important that such political and regulatory uncertainty be kept to a minimum. Further investment in transmission capacity is also likely to be warranted, especially in light of increasing concentration of the Nordic electricity industry. A low level of concentration may have been the most important factor underlying the success of the regulatory reform so far, and it would be a pity if mergers, horizontally and vertically, were allowed to undermine the performance of the industry. There is reason to believe that the difference in end-user prices between Norway and Sweden may be explained by a lack of competition in the Swedish retail market resulting from a combination of vertical integration between generation and retail and high levels of concentration. Similarly, tighter market conditions and more frequent occurrence of bottlenecks in the transmission system are signs of increasing segmentation of the market that may lead to higher prices also at the wholesale level. Only a combination of adequate investment incentives and strict competition policy can ensure the continued success of the Nordic electricity market. In conclusion, it seems fair to say that the electricity market reform in the Nordic countries has been relatively successful in comparison with other electricity markets, for example the California market that collapsed when exposed to severe shocks in 2000–2001. There seems to be four main factors explaining this (see Amundsen, 2005; Bergman, 2005; Amundsen and Bergman, 2006). Firstly, the market design of the Nordic market is simple but sound and to a large extent made possible by the large share of hydropower. Secondly, dilution of market power, attained by the integration of the four national markets into a single Nordic market, has been rather successful. Thirdly, there has been a strong political support for a market-based electricity supply system without intervention in the market mechanisms in stressful situations. Fourthly, the Nordic power industry seems to have a strong voluntary informal commitment to public service. Clearly, only the second and third of the above-mentioned factors are “transferable” as suggestions to other countries whishing to reform their electricity markets, while the first and the fourth to a large extent are country specific. Hence, the experiences of the Nordic electricity market suggest that a market for electricity works well if there are no price regulations and constraints on the development of financial markets and that there is continued political support for a market-based electricity supply system also when electricity is scarce and prices are high. References Amundsen, E.S. (2005). Smooth adjustment vs. melt-down: experiences from a Nordic supply shock and comparison with the California case. Paper Prepared for the EPRI International Conference Global Electri-city Industry Restructuring: In Search of Alternative Pathways, San Francisco, California, May 11–12, 2005.

170

Electricity Market Reform

Amundsen, E.S. and Bergman, L. (2006). Why has the Nordic electricity market worked so well? Utilities Policy (forthcoming). Amundsen, E.S. and Mortensen, J.B. (2001). The Danish Green Certificate System: some simple analytical results. Energy Economics, 23, 489–509, with Erratum in Energy Economics, 24, 523–524. Amundsen, E.S. and Nese, G. (2004). Market power in interactive environmental and energy markets: The case of Green Certificates. SNF-Report 10/2004. Samfunns- og næringslivsforskning AS, Bergen. Amundsen, E.S., Nesse, A. and Tjøtta, S. (1999). Deregulation of the Nordic power market and environmental policy. Energy Economics, 21, 417–434. Aune, F.R. (2003). Fremskrivning for kraftmarkedet til 2020 – virkninger av utenlandskabler og fremskyndet gasskraftutbygging. Reports No. 2003/11. Statistics Norway. Bergman, L. (2005). Why has the Nordic electricity marked worked so well? Paper Prepared for the EPRI International Conference Global Electricity Industry Restructuring: In Search of Alternative Pathways, San Francisco, California, May 11–12, 2005. Bergman, L. and Radetzki, M. (2003). The Continue Project: Global Climate Policy and Implications for the Energy Sector in a Small Open Economy: The Case of Sweden. Multi-Science Publishing Company Ltd, Stockholm. Bergman, L., Brunekreeft, G., Doyle,C., von der Fehr, N.-H.M., Newbery, D.M., Pollitt, M. and Regibeau, P. (1999). A European Market for Electricity? Monitoring European Dergulation 2. Centre for Economic Policy Research, London/Stockholm. Bjørndal, M. and Jørnsten, K. (2001). Koordinering av nordiske systemoperatører i kraftmarkedet – gevinster ved bedret kapasitetsutnyttelse og mer fleksibel prisområdeinndeling. SNF-Report No. 29/01. Samfunns- og næringslivsforskning AS, Bergen. Bye, T. (2003). A Nordic energy market under stress. Economic Surveys, 4, 26–37. Statistics Norway. Bye, T. and Bergh, P.M. (2003). Utviklingen i energiforbruket i Norge 2002–2003. Reports No. 2003/19. Statistics Norway. Bye, T. and Larsson (2003). Lønnsomhet ved tilbakesalg av kraft fra kraftintensiv industri i et anstrengt kraftmarked. Økonomisk Forum, Vol. 57, 26–29. Bye, T., von der Fehr, N.-H.M., Riis, C. and Sørgard, L. (2003a). Kraft og makt – en analyse av konkurranseforholdene i kraftmarkedet. Report of an Expert Group Appointed by the Norwegian Ministry of Labour and Administration. Bye, T., Hansen, P.V. and Aune, F.R. (2003b). Utviklingen i energimarkedet i Norden 2002–3. Reports No. 2003/21. Statistics Norway. Hauch, J. (2003). Electricity trade and CO2 emission reductions in the Nordic countries. Energy Economics, 25, 509–526. Hindsberger, M., Nybroe, M., Ravn, H. and Schmidt, R. (2003). Co-existence of electricity, TEP and TGC markets in the Baltic Sea Region. Energy Policy, 31, 85–96. IEA (2000). CO2 emissions from fuel combustion 1971–1998. IEA Statistics, ISBN 92-64-08506-8. Nordel (2001). The transit solution in the Nordi electricity power system. Feature article in 2001 Annual Report. Nordel (2003). The future infrastructure of the Nordic countries. Feature article in 2003 Annual report. Treckin (2004). The one-stop information network for tradable renewable certificates, http://www. treckin.com Unger, T. and Ahlgren, E.O. (2003). Impacts of a common green certificate market on electricity and CO2 emission markets in the Nordic countries. In T. Unger (ed.), Common Energy and Climate Strategies for the Nordic Countries – A Model Analysis. Ph.D. dissertation, University of Gothenburg, Gothenburg, Sweden. Unger, T. and Alm, L. (2000). Electricity and emission-permits trade as a means of curbing CO2 emissions in the Nordic countries. Integrated Assessment, 1, 229–240. von der Fehr, N.-H.M., Amundsen, E.S. and Bergman, L. (2005). The Nordic Market. Signs of stress? The Energy Journal, 61–88, European Electricity Liberalization Special Issue. von der Fehr, N.-H.M., Hagen, K.P. and Hope, E. (2002). Nettregulering. SNF-Report No. 29/01. Samfunns- og næringslivsforskning AS, Bergen.

PART III Evolving Markets

This page intentionally left blank

Chapter 6 The Electricity Industry in Australia: Problems Along the Way to a National Electricity Market ALAN MORAN1 Institute of Public Affairs, Melbourne, Australia

Summary2 In Australia prior to 1994 virtually all electricity was supplied through vertically integrated state monopolies. A decade later, the integrated monopolies had been disaggregated into different businesses with the competitive aspects of supply (generation and retailing) reconstituted into dozens of independent firms, many of them privately owned and the rest “corporatized” and operating at arms length from their government owners. Monopoly aspects of supply are regulated by agencies independent from the jurisdictional governments. The key milestones have been as follows: Sequence of events: ● ●

●

●

● ●

●

Industry Commission Report into electricity, 1991. Institute of Public Affairs/Tasman Institute report on Victorian electricity corporatization and privatization, 1991. National Grid Management Council’s National Electricity Market (NEM) Paper Trial, 1993/1994. Victorian Electricity Market was commenced on 1994. This comprised six major generation businesses, a transmission business and five distributor/retailers. Distribution regulated by independent body. Market gradually opened to competition in 1995–2001. Some price caps remain on household supply. Victorian electricity and gas privatizations 1995–1999. Competition Principles Agreement, 1995 (A$4.2 billion of Commonwealth funds were set aside for the period of 2005/2006 to implement electricity reform). New South Wales (NSW) Electricity Market, 1996. Competitive arrangements were established that were similar to Victoria’s. Market opened to competition in 1996–2002 with some price controls on households remaining.

1

Helpful comments were received from many people including Perry Sioshansi, Paul Simshouser (Braemar Power), Darren Barlow (Ergon Energy) and Ben Skinner (TRUenergy). 2 Prices here are quoted in $A or cents, which refer to Australian cents. The Australian dollar is worth around 75 US cents.

173

174

● ●

●

● ●

●

Electricity Market Reform

Agreement on National Electricity Code, 1996. Start of NEM, 1998 with the National Electricity Code Manager and the Australian Competition and Consumer Commission (ACCC) setting rule changes at the national level and ACCC setting transmission prices. Queensland Electricity Market, 1998. Market opened to competition in 1998–2004 except for households. South Australia privatization, 2001. Full retail competition phased in 1998–2003. Australian Energy Regulator (AER) (price setting) and Australian Energy Market Commission (AEMC) (Rule changes) commenced operation in 2005. Western Australia de-aggregation of supply but with the generation left within one business, 2005.

Outcomes of these developments have been reductions in prices, especially for commercial users, which were previously subject to Ramsey-type price gouging.3 The reformed system has delivered increases in capacity in line with market needs and vast improvements in productivity and reliability across the industry. Government interventions have, however, not been eliminated and continue to threaten the on-going orderly development of the market. Among these potential market distortions are the ramifications stemming from state governments’ ownership of over half of the industry. Although all government businesses are corporatized and operate under company law, government ownership brings corporate inflexibilities and sometimes means political interference in key commercial decisions. More generally, electricity remains an industry with a high political profile. Federal and state governments see electricity supply and pricing as providing them a somewhat unique legitimacy to control. At the very least, this brings distractions to the industry’s entrepreneurship and there is ample risk of more serious consequences to the industry’s efficiency.

6.1. The Market Breakdown and Supply Profile 6.1.1. The market profile Australian electricity demand is about 200,000 GWh/annum, which is transmitted along some 850,000 km of lines of which some 26,000 km is along circuits rated at 220 KV or more. There are some nine million customers with demand being divided almost equally between households and industry. Figure 6.1 illustrates this. The geographic spread of generation facilities and transmission and distribution lines is illustrated below.

6.1.2. The fuel profile of Australian electricity generation Australia is fundamentally a coal-based electricity system. Coal represents some 85% of electricity supply, roughly one third of which is Victorian and South Australian brown coal.

3

Under the state owned vertically integrated monopolies, large, footloose users negotiated cost-based prices (below cost in the case of some aluminum smelting contracts). Household supply benefited from a cross-subsidy paid for in higher prices to most business customers. Since the market has been operating almost all business customers and many household customers have seen real price reductions.

175

The Electricity Industry in Australia Electricity consumption by sector 1.1%

1.5%

23.8% 46.9% 26.7%

Industry Residential Transport

Commercial Agriculture

Fig. 6.1. Electricity consumption. Source: NEMMCO.

Hydro capacity is relatively small despite the large land area and is nearly fully developed. Tasmania has further minor potential but green activism will prevent any substantial new development. The share of hydro within total supply has, therefore, been falling to its current level of about 7%. Gas has shown a modest increase growing from less than 2% of generation 15 years ago to 7–8% at present. Much of this is in the least heavily populated states of South Australia, Western Australia and the Northern Territory where coal is more expensive. Elsewhere gas mainly fills a peaking role due to its lower capital but higher fuel cost. Figure 6.2 illustrates the fuel sources of electricity. Australia’s eight States and Territories vary in size, with New South Wales and Victoria being the most populous and the two Territories, the city-state Australian Capital Territory and the vast Northern Territory each having populations of a few hundred thousand. Most people live along some 4000 km of coastline from Adelaide in South Australia to Cairns in North Queensland. This has resulted in a long, skinny interconnected grid: the NEM. The

176

Electricity Market Reform

island of Tasmania is in the process of connecting to the NEM via a 600 MW undersea cable. The only other significant population center, South West Western Australia, has its own grid, which is not economic to connect to the NEM. The same applies for smaller isolated population centers such as the Northern Territory. Each state has only one pricing node4 that sets price for all customers and generators within that state. Generally there is little transmission congestion within the states, and a moderate amount between the states – with “interconnectors” constraining ⬃5% of the time. At these times, the regional prices diverge, sometimes by thousands of dollars per MWh, but very rarely for more than a few hours. Queensland and Victoria are major net exporters of electricity via the NEM (to South Australia and New South Wales respectively) and New South Wales is also an exporter because it hosts the jointly owned Snowy Mountains Hydro-Electric facility. Tasmania is hydro based and Basslink will allow Tasmania to export its hydro at times of high mainland prices and import baseload low-priced mainland-generated electricity at other times. Figure 6.3 illustrates the relative size of the NEM States in terms of energy sent out.

Generation by fuel source 250

GWh

200 150

Gas Coal Hydro

100 50 0 1991

1993

1995

1997

1999

2001

2003

Fig. 6.2. Source: ESAA.

Total energy sent out 2003–2004 2.5% 6.5%

28%

27%

36%

QLD NSW TAS

VIC SA

Fig. 6.3. Source: NEMMCO.

4

NSW has a second pricing node in the snowy mountains scheme, however there are no significant customers at this node.

The Electricity Industry in Australia

177

6.2. Reforming the State Owned Integrated Utilities 6.2.1. The process of reform The development of a market for electricity got underway in the early 1990s and had three precursors: ●

●

●

the recognition that other countries were achieving considerably greater efficiencies than Australia in electricity supply;5,6 National Competition Policy (NCP) involving a general review of the operations of “essential facilities” (which were, in the main, owned by governments) and a requirement that they be opened to non-affiliates on reasonable terms; and the consequences of poor financial circumstances in the States of Victoria and South Australia resulting in new governments which sold its energy assets partly in pursuit of a privatization agenda and in part to reduce debt.

The State of Victoria initiated the reform process. By the early 1990s the then Labor Government had commenced a process of reform particularly focusing on labor shedding in the monopoly supplier, the State Electricity Commission of Victoria (SECV). A Liberal (Conservative) Government, which won office in 1992, set about a much more aggressive reform and privatization process. That Government had a strong philosophical belief in the beneficial effects that capital markets could bring to a business as well as being attracted to the idea of transferring risk to private equity. However, the Government was equally determined to get the structure of the market right first. In this, lessons were learned from the outcome of the structural weaknesses in the UK wholesale generation market, where the two dominant suppliers were able to operate so that prices remained high (See Newbery). Great care was therefore taken to establish competition in supply with this being given a higher priority than maximizing sale proceeds. 7500 MW of plant was privatized as seven generating companies with transitional prohibitions on re-aggregation. For downstream supply, five distribution/retail businesses were created and transmission assets were vested in a single business. An independent regulatory framework, the Office of the Regulator General (ORR) later renamed the Essential Services Commission (ESC) was put into place to set prices on the monopoly network assets. This body also gathered performance data that would prove invaluable in defusing claims that privatization had led to higher prices and lower reliability. The program for reform was a challenging task. Not only did it involve the need to establish a regulatory framework, in which Australia had no experience, but also because the reform process faced considerable hostility from the unions, the Labor opposition and much of the media. However by the time of the Liberal Government’s fall in 1999, the electricity industry had been privatized and assets transferred to seven generation businesses, five distributor/retailers plus a transmission business. There has since been some merger rationalization as well as some new entry. The sale process earned A$23 billion, far more than the $9–10 billion that was widely expected. The gas industry’s downstream assets were also privatized; gas production was always privately owned. 5

Project Victoria: A Rebuilding Strategy for Electricity in Victoria, Tasman Institute/Institute of Public Affairs, 1991. 6 Industry Commission, Energy Generation and Distribution Report, No. 11, 1991, http://www.pc.gov.au/ic/ inquiry/11energy/finalreport/index.html

178

Electricity Market Reform

The fully privatized Victorian electricity and gas industry has three main retailers (CLP Power’s TRUenergy, Origin Energy, and AGL), four distribution businesses (CKI-Powercor, Alinta, AGL, and SPI), a single transmission business (SPI) and five major generators (International Power, CLP Power’s TRUenergy, Loy Yang Power, Ecogen, and Southern Hydro). There have also been significant new entrants into the competitive generation and retailing businesses. South Australia, which is only one quarter Victoria’s size in terms of customers, privatized its industry (in 1999 and 2000). Its previous government monopoly supplier had been the ETSA Corporation. As in other jurisdictions, ETSA’s disaggregation preceded the industry’s full privatization. There is now one distributor, the CKI owned ETSA Utilities and one transmission company, ElectraNet. Only one retailer was created at privatization, which AGL bought, however there has been significant new entry. Privatized generation is owned by International Power, NRG, and TRUenergy. AGL and Origin built significant new generation capacity early this decade. In Queensland, the Queensland Electricity Commission (QEC) was a vertically integrated organization responsible for virtually all electricity supply and was disaggregated in several steps into separate distributor/retailers, a single generation business and an SPI. In 1997 a further progressive disaggregation took place which has resulted in the present industry profile comprising government ownership of two main distributor/retailers (Energex and Ergon), a transmission business (Powerlink) and four generation businesses (Enertrade, CS Energy, Stanwell, and Tarong Energy). There is also private ownership of generation including Intergen and NRG. The New South Wales Government’s prior ownership of the integrated Electricity Commission has undergone several iterations of disaggregation but the separate businesses remain under government ownership. Individual suppliers include four distributor/retailers (Energy Australia, Integral, Country Energy, and Australian Inland) a transmission business (Transgrid) and three coal-based generation businesses (Delta Electricity, Macquarie Generation, and Eraring Energy). There are also two smaller private generators and the Snowy Hydro jointly owned with the Victorian and Commonwealth Governments. The Commonwealth, NSW and victorian governments, which jointly own Snowy, announced in early 2006 that they are to privatize it. The other jurisdictions have also disaggregated their formerly integrated industry but Tasmania (which will be linked to the mainland grid in 2006) and Western Australia have retained government ownership. The major ESI businesses and their ownership are shown in Table 6.1. Overall Australia’s ESI remains mainly under government ownership as illustrated in Figure 6.4. 6.2.2. The national setting for electricity market reform The Victorian disaggregation and privatization took place as a result of financial pressures and pro-competition/privatization views that were especially prevalent in that state. In the case of other states, the reforms followed on a program of NCP, agreed by the state and Commonwealth Governments in 1996. This was given expression in a new provision of the Trade Practices Act, Part IIIA, an important provision of which required essential facilities to be opened to competition, clear structural separation for the competitive and monopoly parts of integrated businesses and, where practicable, non-natural suppliers were to be divided into rival businesses. The competition policy reforms called for natural monopolies to be opened to all users on terms that were fair and reasonable. Provisions were made to ensure that regulatory

179

The Electricity Industry in Australia Table 6.1. Major Australian electricity supply businesses. Ownership

State of operations

Major generators Macquarie Delta Snowy Eraring TRUenergy CS energy Loy Yang Intergen Tarong NRG flinders International power Stanwell Origin Ecogen NRG Gladstone Hydro Tasmania Southern hydro Enertrade

NSW government NSW government NSW, Vic federal governments NSW government CLP Qld government/Intergen AGL, Tokyo electric and others International consortium Qld government NRG International power Qld government Australian private International consortium NRG, Comalco Tas government Meridian energy Qld Government

NSW NSW NSW, Vic NSW Vic, SA Qld Vic Qld Qld SA SA, Vic Qld Vic, SA, Qld Vic Qld Tas Vic Qld

Major transmission SPI Transgrid Powerlink Electranet

Singapore power NSW government Qld government public private consortium

Vic NSW Qld SA

Major distribution AGL Alinta Aurora Citipower/Powercor/ETSA Country Energex Ergon energyAustralia Integral energy SPI

Australian private Australian private Tas government CKI NSW government Qld government Qld government NSW government NSW government Singapore power

Vic, SA Vic, WA Tas Vic, SA NSW Qld Qld NSW NSW Vic

Major retailers AGL Aurora energyAustralia Integral energy Country Energex Ergon Origin TRUenergy

Australian private Tas government NSW government NSW government NSW government Qld government Qld government Australian private CLP

Vic, SA Tas NSW NSW NSW Qld Qld SA, Vic Vic SA

180

Electricity Market Reform

Ownership of NEM assets

36%

64%

43%

Generation

57%

Transmission Public

50%

50%

Distribution

55%

45%

Retail

Private

Fig. 6.4. Source: ESAA

arrangements were put in place to determine where such access was required and, in the event of disputes, the prices at which the access was to be made available. Far and away the most important monopolies over essential services were those of governments themselves – in addition to electricity and gas they included ports, airports, rail, and telecommunications. Indeed, the government owned or controlled businesses were the only ones that in practice had the integration and monopoly characteristics that required the regulatory control.7 6.2.3. Market governance The national market was devised jointly by the state and Commonwealth governments and refined and approved by the national regulator, the ACCC. It originally envisaged two bodies: the National Electricity Code Administrator (NECA) and the National Electricity Market Management Company (NEMMCO) being responsible respectively for the market rules and the market scheduling and planning matters. In the event, the confluence of the electricity market developments and NCP meant that the national regulatory body, the ACCC had to have a role in approving market rule changes because of the inherent monopolistic “collusion” that the NEM entails. The ACCC also was given the role in setting prices for the transmission lines and for gas transmission and the market code for gas. In addition, state regulators were put in place to set prices for the distribution lines so that in all there were some 12 bodies involved in economic regulation of the industry. Also, there were Ministerial Councils that set agenda issues and generally sought to influence the market. Fortunately, inconsistencies in decisions from such a plethora of regulatory bodies was not as serious as it might have been since there were strong liaison relationships forged through an informal “Energy Regulators’ Forum”. Even so, the process of National Electricity Code change was proving somewhat unwieldy with each proposal being examined first by NECA and then by the ACCC. New arrangements were agreed in 2005 that introduce a rationalized control of the industry so that the AEMC is responsible for market rules for both gas and electricity and the AER is responsible for policing the rules and for network price setting at all levels of supply. NEMMCO’s functions (as Market/System Operator) remain as those of the ACCC, which also has common staff with the AEMC and AER, with regard to control of mergers and monopolies. It is, of

7

Exceptions included gas where AGL had a monopoly in NSW and, arguably, BHP steel, the integrated plant of which had been found by the High Court, in the Queensland Wire case, to be required to be opened to competition. (See W. Pengilly www.deakin.edu.au/buslaw/aef/publications/workingpapers/ swp2004_181.pdf)

The Electricity Industry in Australia

181

course, uncertain that the changed arrangements will prove more workable especially as there are statutory requirements for these bodies to follow the policies established by the collective state and federal governments. The AEMC is responsible for rule making and market development. The rule-making role does not involve initiating changes to the Rules other than where the change involves correcting minor errors or where the change is of a non-material nature. Rather, the role involves managing the rule change process, and consulting and deciding on rule changes proposed by others. In regard to its market development function, the AEMC conducts reviews at the request of the Ministerial Council on Energy or at its own volition on the operation and effectiveness of the Rules or any matter relating to them. In doing this, the AEMC relies on the assistance and cooperation of industry relationships and interested parties in its decision making. 6.2.4. Structural developments At the onset of both the Victorian and the later National Market rules, the provision implementing structural separation of generation, transmission and distribution/retailing contained no specific long-term measures to prevent re-aggregation. This was because there were no firm views as to the most productive structure of the industry, only that the previous state owned integrated monopolies were not optimal. Although retailing and distribution were sold as combined units, they were to be “ring fenced” to prevent the distribution business favoring its affiliate. The ring fencing has generally proved satisfactory. Not only have incumbent affiliates not been favored but new retailers have entered the market and all five of the original Victorian host distribution business/retailers now have separate companies handling the two activities.8 This reflects the very different types of business involved. Retailing involves strong marketing and risk management skills in assembling and promoting packages of supply, while distribution is much more concerned with maintaining and reducing costs of an established business. Two of the three major retailers, Origin and TRUenergy, no longer are affiliated with a distributor. Their disaggregation was driven by a search for improved shareholder value whereby different types of business can appeal to different types of investor. There has also been a trend towards an unexpected form of re-aggregation. This has not been to restore monopolistic supply, which in any event would be combated by general anti-trust laws administered by the ACCC. Instead, all the privatized retailers and many of those remaining in government hands have moved to acquire some generation of their own. Also, Snowy Hydro and several privatized generators have created retail arms. These developments are a function of the need that generators and retailers see to manage risk. Risk management strategies that firms have employed, cover a gamut from different forms of contracting through ownership of supply. Seeking some control over their supply (for retailers) or customers (for generators), is a risk management strategy that follows from the wholesale price shifts to which electricity is now subject. Early concerns that retailer and generator amalgamations would lead to monopoly power and market inefficiencies have tended to abate as a result of evidence that competition is bringing lower prices and retail churn. The generator–retailer re-aggregation that has occurred has not been anti-competitive. The fact is that electricity retailers like other firms 8

Mergers have reduced the original five host distributors/retailers to three retailers and four distributors. Only one of the retailers retains common ownership with its “host” distribution business.

182

Electricity Market Reform

that assemble supplies sourced from affiliated and non-affiliated businesses are forced by market circumstances to ensure that the affiliates are not favored. The risk of high-price occurrences are too great for electricity retailers to gamble on fulfilling most of their needs in-house and the costs of alienating other sources by unprofessional behavior far outweigh any short-term benefits possible through collusion with in-house suppliers.

6.2.5. Retail price control and competitive churn Retail was a part of the electricity industry envisaged as being contestable and requiring no more regulation than is found in other retail activities. Given its relatively small share of aggregate revenue, retail was also envisaged as being only a minor actor within electricity supply. However, retailing’s importance has been re-assessed. In a competitive situation retail, as the interface with the consumer, drives efficiencies by signaling demand shifts and eroding cross-subsidies. For businesses, full retail competition has been extended to all but the smallest customers and in the four major eastern states probably half the commercial load has shifted from its original retailer. Governments have been more cautious about deregulating household supply. In NSW, South Australia and Victoria, retail competition at the household level has been accompanied by maximum prices that make it less attractive for retailers to poach customers. Nonetheless there has been a quite considerable churn rate – some 16% in NSW, 42% in South Australia and 44% in Victoria after 3 years of open competition.9,10 In Queensland, Tasmania and Western Australia households are captive to their host retailers which are all owned by the respective state government. In Queensland, household rates are fixed and managed so that Brisbane subsidises the provincial areas. The Brisbane-based supplier is levied a surcharge which is passed to the supplier of provincial areas. The crosssubsidy would be placed under considerable pressure with full retail competition (See Haas in this volume) and the electoral implications of this present the main barrier to the Queensland Government allowing full retail competition. In September 2005, the Queensland Government announced that full retail competition, albeit with a regulated price cap, would be introduced in 2007. Table 6.2 shows the timetable for retail competition by state and user tranche. 6.2.6. Other forms of retail regulation On top of price safety nets, the Labor state governments have all imposed their social and green policy objectives via retail regulations for domestic customers. This has resulted in a considerable mish-mash of compliance requirements for retailers selling to small customers and reduced the potential for competition. From the social policy side we have seen prohibitions on pre-payment meters and latepayment fees supposedly in the interest of protecting the poor. 9

For Vic and NSW, http://www.nemmco.com.au/data/ret_transfer_data.htm. For S.A. http://www.saiir. sa.gov.au/site/page.cfm?u⫽4&c⫽1435 10 These figures refer to small customers, some 95% of which are households. They also include some double counting where customers have changed retailers more than once. An alternative measure is of customers who are no longer with their original retailer. For Vic as at September 2005, this is 32.5% and for NSW 11%; these latter figures exclude customers who have new supply contracts with their original retailer.

183

The Electricity Industry in Australia Table 6.2. Timetable for retail competition. Date for eligibility

Site thresholds

Estimated number of customers

Percentage of total energy

New South Wales

October 1996 April 1997 July 1997 July 1998 January 2001 July 2001 January 2002

⭓40 GWh ⭓4 GWh ⭓750 MWh ⭓160 MWh ⭓100 MWh ⭓40 MWh All sites

47 660 3560 10,860 19,000 53,000 3,000,000

14 29 40 47 49 53 100

Victoria

November 1994 July 1995 July 1996 July 1998 January 2001 January 2002

⭓5 MW ⭓1 MW ⭓750 MWh ⭓160 MWh ⭓40 MWh All sites

47 380 1900 6900 N/A 2,100,000

23 29 41 49 N/A 100

Queensland

March 1998 January 1999 January 2000 1 July 2004

⭓40 GWh ⭓4 GWh ⭓200 MWh ⭓100 MWh

80 540 8110 7890

16.0 15.0 15.0 8.0

South Australia

December 1998 July 1999 January 2000 January 2003

⭓4 GWh ⭓750 MWh ⭓160 MWh All sites

160 760 3360 720,000

30.0 40.0 50.0 100.0

Western Australia

July 1997 July 1998 January 2000 July 2001 January 2003 January 2005

10 MW 5 MW 1 MW 230 KW 34 KW 50 MW

N/A N/A 120 450 2550 10,000

N/A N/A N/A N/A N/A N/A

Tasmania

July 2006 July 2007 July 2008 July 2009 July 2010

20 GWh 4 GWh 750 MWh 150 MWh All customers

10 54 295 1030 230,000

N/A N/A N/A N/A N/A

From the green side we have seen irksome rules like requiring minimum area allocations on retailers’ bills for graphs on usage and implied carbon release. More significant are the requirements on retailers to source different percentages of variously defined green power within their aggregate supply. This is addressed in Section 6.7. 6.2.7. Control of transmission Each State has one transmission owner but the planning models are inconsistent. In Queensland, NSW and Tasmania, the transmission owner itself is responsible for planning. In Victoria, a government agency, VENCorp, is responsible for planning the network and SPI owns and operates the assets, whilst South Australia has a hybrid model. SPI in Victoria is unique in that it acquired assets formerly owned by TXU so that it now also owns part of Victoria’s distribution network.

184

Electricity Market Reform

A centrally planned provision of transmission was the basic model adopted from the outset. However, it was also recognized that transmission and new generation are alternatives. If transmission is provided free or at regulated prices this may discourage a more rational and lower cost development of new generation. The trade-off between nearby and remote generation (via transmission) is uniquely critical for Australia, where distances between load centers and therefore the cost of transmission are very large, and fossil fuel sources are relatively inexpensive and quite widespread. This led to provision being made for entrepreneurial interconnects in the National Electricity Law. And Transenergie, a subsidiary of Hydro Quebec, built two of these entrepreneurial links where transmission links were less than robust. Transenergie sought to finance these links by selling generators’ access rights to markets and by itself arbitraging price differentials. This merchant transmission gave rise to issues concerning the circumstances under which a regulated augmentation of links should be permitted. A long series of hearings on a regulated link between South Australia and New South Wales resulted in stalemate, with the NSW government transmission business (Transgrid) apparently abandoning its proposal, possibly because NSW is not envisaged to have a generation surplus in future years.11 In the event, the merchant links in Australia could not compete against the links receiving a regulated return and have applied for and been given regulated status.12 The danger is that links which are financed by a compulsory charge on the customer, might lead to incentives to site generation in places that are distant from major markets. If someone else is paying for transmission, the rational generation business will be indifferent to its costs thus distorting the efficient trade-off transmission costs and generation costs. It is argued that the externalities are too great to allow profitable merchant transmission since the benefits of lower prices (actually arbitraged prices) accrue to all and not only to those paying for the asset. However, this is not markedly different from the situation concerning a new generation facility, which will tend to suppress the price of all delivered electricity in its interconnected region. Few would argue that by analogy all generation should therefore be government owned or subsidized even though many argue for a form of general overhead support in the form of capacity payments. The fact is that supply across the economy is seldom unaccompanied by some externalities. Associated with the claim that transmission would be inadequately provided in the absent of it being made subject to regulated support, is the contention that a transmission 11

This has brought a voluminous level of studies. Those in Australia include the skeptical like Mountain, B. and Swier, G. Entrepreneurial interconnectors and transmission planning in Australia, The Electricity Journal, March 2003. London Economics in its work for the ACCC (Review of Australian Transmission Pricing, 1999), also concluded that entrepreneurial links could not cover their fixed costs. This skepticism is also seen in the work of Joskow and Tirole (e.g. Merchant Transmission Investment, CMI Working Paper, 24 The Cambridge-MIT Institute, 2003). The Australian 2002 Paper on Independent Review of Energy Market Directions (www.energymarket review.org) saw a possible role. Littlechild has been more supportive both in studies in Australia and Argentina (e.g. Littlechild, S. (2004) “Regulated and merchant interconnectors in Australia: SNI and Murraylink revisited.” Applied Economics Department and The Cambridge-MIT Institute, Cambridge University, Cambridge Working Papers in Economics CWPE No. 0410 and CMI Working Paper 37; and Stephen C. Littlechild and Carlos J. Skerk Regulation of transmission expansion in Argentina CMI, Working Paper 61, University of Cambridge, Department of Applied Economics, 15 November 2004). Further complications to the issue also result from the flowgate and nodal pricing debate. 12 This tends to confirm a dimension of Hogan’s “slippery slope” hypotheses whereby due to its ability to compel payment government ownership tends to drive out private ownership. Hogan W.W., Transmission Market Design; http://ksghome.harvard.edu/⬃whogan/trans_mkt_design_040403.pdf

The Electricity Industry in Australia

185

line has market power and its prices should be regulated. However, for the most part, transmission inter-ties or inter-connects offer no more market power than that of a significant generator portfolio. Inter-ties in Australia can account for some 35% of supply (Victoria to South Australia) but normally provide much less than this. Their market power is confined to influence over those wishing to export and such firms are normally capable of writing contracts to cover and vulnerabilities they foresee. Issues on how best to allow expansion of transmission, especially in terms of the regional linkages has been subject to heated debate in Australia. An uneasy compromise is presently in place for transmission under which regulated links will be permitted as long as a net market benefit is judged by the regulator to be the outcome and as long as the proposed link is the best of a range of feasible alternatives. This, however, remains dissimilar from the decision making structure that is seen in the generation sector or in markets more generally since it may incorporate some to the network benefit externalities which an comparable investment in a new generator would not capture. The competing solutions that generation and transmission often offer mean disputes about the merits of a new transmission solution are likely to remain. These may be exacerbated since the Queensland Government has encouraged new government owned capacity to be built in that State, driving down prices below those in NSW and is seeking to augment transmission links. Other states regard this as facilitating dumping and are opposing to having expanded capacity financed as a regulated link since most of the costs fall directly on consumers. These considerations have been further complicated by the growth of subsidized wind generation. Wind power is always likely to be relatively dispersed and remote and, in addition to production subsidies, its sponsors have already extracted concessions from some governments that smear its transmission costs. Over a thousand MW are planned in South Australia where conventional capacity is only 3000 MW. 6.2.8. Network price setting The ACCC is responsible for setting prices for electricity transmission lines (as well as those of gas). Local State regulators at the present time are responsible for setting distribution prices. The price setting process has assumed a vast complexity as the regulator and the businesses each hire accountancy and economics advice to determine the appropriate prices. Regulated businesses are never likely to express satisfaction with the determinations of a regulator, but for the most part over recent years the outcomes have been more predictable and less contentious. There remains the risk that price cuts can induce sub-optimal investment. In Victoria, following a price reduction on distribution businesses averaging 15% in 2001 further real price cuts averaging 14–26% are proposed for 2006 when the businesses claim that all the fat was cut out in the first price re-set. In the case of gas transmission emerging competition is generally to be seen across the country. In spite of this, the ACCC has been seeking to maintain its regulatory powers and its decisions have been at variance with those of the market itself. Both the Minister for Energy,13 who has review powers over certain ACCC decisions, and the Productivity Commission,14 the advice of which has been sought by the government, have suggested 13

Macfarlane I (Minister for Industry, Tourism and Resources) (2003), Applications for Revocation of Coverage of Certain Portions of the Moomba to Sydney Pipeline System: Statement of Reasons, www.industry.gov.au/ content/itrinternet/cmscontent.cfm?objectid⫽F14AF2D8-0F2F-FF46-E371BA8885EA4888& indexPages⫽/content/whatsnew.cfm&CFID⫽30998&CFTOKEN⫽24724739. 14 10 August (2004). Review of the gas access regime, productivity commission, report No. 31.

186

Electricity Market Reform

that the ACCC’s powers be curtailed somewhat where, as is evidenced with gas, competition is itself providing adequate market disciplines. The danger of a regulator setting transmission prices too low is that this will result in inadequate new investment. Setting prices too low on existing network facilities (transmission and distribution), irrespective of any general provisions that are in place to maintain quality, will also result in inadequate maintenance. A claimed outcome of mandated price cuts by the Queensland regulator was a series of blackouts in 2004. On the other hand is the risk of setting regulated transmission returns too high. Where these are provided by compulsory charges there is likely to be over-building with the previously discussed adverse impacts on alternative, more efficient solutions like additional local generation capacity. This has been foreseen in the NEM Code which has provisions for rationalizing redundant facilities. These however are difficult to activate.

6.3. Operations of the National Market 6.3.1. The spot and contract markets The Australian National Market is underpinned by a “gross pool” system under which all major suppliers must bid. There is no capacity payment system and the market for energy is linked and simultaneously cleared to eight separate markets for reactive power and other “ancillary services”. Although virtually all power must be bid into the pool (a major exception being wind power, which must be taken as it comes) few customers or suppliers would find it prudent to rely solely on pool prices. In effect the market is now a multiplicity of bilateral contracts between generators and retailers, usually in the form of contracts for difference, strongly underpinned by the “gross pool”. Essentially the NEM is a one-way market with the demand emerging from whatever end-users require and with supply being offered by generators at different price levels. Through the pool, all supply is paid the same price, that of the highest bid supply that is dispatched. In the short term, generators are relatively indifferent to the price of that part of their supply for which they have contracts (probably 90% plus of the market) and will usually bid close to their marginal costs for this part of the load.15 Compared to a price norm of about $A30/MWh, at present prices can rise to as much as $A10,000/MWh which makes retailers especially keen to be fully contracted or to manage spot exposure through ownership of peaking capacity. A complex National Electricity Code controls the rules under which generation and transmission are placed on the market. Generators bid in a maximum of 10 price bands and although there are restraints on price re-bidding, there are none on shifting energy between different price bands meaning, in effect, that prices can be changed at any time prior to 30 seconds of dispatch. Each generator has controls built into the dispatch algorithm covering ramp rates and minimum loads. Loads can also bid to be offloaded though few do. In some cases this is due to lack of “smart” metering though it is likely that only those users, for example smelters, who have 15

For a discussion on whether the generators exercise market power see, Moran, “Is there market power in Australian electricity generation?” ACCC Regulatory Conference, July 2004, http://www.ipa. org.au/ files/ammarketpower.pdf. For an analysis of rational bidding strategies see Peter Cramton, Competitive bidding behavior in uniform-price markets. Hawaii International Conference on System Sciences, January 2004.

The Electricity Industry in Australia

187

energy as a very large share of their aggregate costs would consider it worthwhile to make savings by shutting down and re-starting. It is an assumption of those calling for mandatory roll-outs of smart meters that this will markedly influence household behavior. It is, however, likely that any such influence would be in the form of contracts that suppliers arrange with households for automatic partial disconnect for short periods. How prevalent these will be is open to conjecture. As an insurance against inadequate capacity, there is a (supposedly temporary) provision for a “Reserve Trader”. This entails NEMMCO seeking additional supply offers (or demand side offers) in the event of it determining that a shortfall will occur in forward supply. Such a shortfall would be determined from forecasts of maximum capacity 2 years ahead that generators are required to provide, forecasts that are married to NEMMCO’s own demand forecasts. Under current standards, which are a combination of N-1 and unserved energy standards, NEMMCO is required to ensure 850 MW of reserve is carried across the entire NEM – including during the periods of extreme demand – to provide the required level of supply reliability. Though the Reserve Trader provisions have in fact been used, they have an intrinsic deficiency in that they simply call forth supplies that are available in any event. No government body would contemplate building additional capacity solely for reserve purposes. Reserve Trader capacity, therefore simply adds to costs unless the market manager has superior prescience to the various market participants. This is because, absent the government intervention, to the degree that supply shortfalls emerge, prices will be driven up and suppliers have every incentive to make plant available that was formerly mothballed, due for scrapping or being restrained to economize on maintenance. Moreover, a Reserve Trader policy carries seeds that can distort the entire market. If the Reserve Trader is in place and is offering higher prices than those anticipated by market participants, a rational generator will withdraw capacity thus exacerbating the apparent future shortfall. If such actions were to snowball, there would be a progressive increase in the apparent shortfall and an increasing need for the authorities to contract outside of the normal market, thus undermining it.

6.3.2. Market integration: access to transmission Although Australia does not have a nodal pricing system, there are procedures in place that in principal, lead to new regions being created when areas are islanded with significant price separation for more than 40 hours/year. However, political pressure has prevented new regions being created, notably in Queensland where the entire state remains a region in spite of considerable transmission weaknesses between the populous south east of the state and the northern parts. Loss factors for transmission and capacity constraints mean that there may be significant differences in the spot price for any trading interval across NEM regions, though market separation rarely occurs for more than an hour or so. Once a region is created, inter-regional settlements are auctioned. These settlements represent the difference between the value of electricity in the region where it is generated and its value if sold in another region. The settlement residue that accumulates is made available to the market by the conduct of an auction. Holders of this auctioned revenue value have a form of hedge that contributes to facilitating inter-regional trade by providing market participants a risk management mechanism as protection against high prices. The risk is however not eliminated since if the link constrains, revenues are not able to benefit fully from high prices in one region because of volume shortfalls.

188

Electricity Market Reform

6.3.3. Issues concerning generator market power Market supplies in Australia have on the whole been sufficiently disaggregated to prevent monopolistic power to be exercised other than on a transitory basis. In the interconnected NEM of six regions, the largest generation business has only 14% of the market and two other businesses each have over 10%. Concentration is somewhat greater in individual regions and these sometimes have constrained interconnects. Three businesses have 10% or more of the national market’s capacity (one of which, Snowy, is a hydro generator with the maximum annual output equal to about one quarter of capacity). A further 11 businesses each have 2–8% of the capacity market while a host of smaller suppliers collectively account for about 10% of the market (Fig. 6.5). In some states the capacity is more concentrated. Thus, NSW has three baseload generators (all government owned) controlling 70% of capacity and it is claimed that they have used market power on occasion to ensure that the mandatory insurance system for retail prices created by government regulation (and discussed in Section 6.7) does not accumulate income. NSW market shares are shown in Figure 6.6. In addition to this level of market supply, there are two transmission links from Queensland, theoretically rated at a combined 880 MW.

St IP an w e O ll pt im Ec a og en Fl in de rs O th er

D el ta Sn ar o in w g y en Ya er llo gy ur n (C C S LP) en e Lo rgy y ya n Ta g ro ng N R G Er

M

ac q

ua rie

%

Major generators’ market share 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0

Fig. 6.5. Source: NEMMCO.

Production capacity: NSW

Fig. 6.6. Source: NEMMCO.

nt

ry

L ED

ou C

nk

e

ba ed

L AG

g

ta

ar in Er

D

el

y w Sn o

Si th

R

M

ac q ge ua n ri

e

5000 4000 3000 2000 1000 0

189

The Electricity Industry in Australia

In Victoria and South Australia, which have strong transmission capacity links with each other, the market supply includes that which can be transferred through Snowy either from the hydro facility itself or from NSW power suppliers. As from early 2006 a further transmission connection from Tasmania will operate with a capacity of 300 MW into Victoria and 630 MW into Tasmania. Even though CLP has over 30% of the capacity, there are several other powerful suppliers and there are few concerns about abuse of market power (Fig. 6.7). The fact that there is so much flexibility in bidding under the Australian system has contributed to unease in certain quarters that price oscillations have been due to supplier market manipulation. Of course, such price oscillations have always been present – they are inevitable because of the wide and sometimes rapid variations of demand (and occasionally supply). They were masked in the past because the monopoly provider called in higher merit order plant to cover peaks or sudden needs without this actually being specifically priced. The sharp oscillations in price now visible have led to claims that the generators are unfairly “gaming” the market, driving up prices to take advantage of opportunities in which there is a monopoly. A lengthy series of investigations into the structure of the bidding rules was underway between 1999 and 2002. Similar debates were held in the UK and other countries. In the event there was recognition that temporary market power is common to many industries and that attempts to combat it by fixing or constraining prices could exacerbate the underlying conditions that create it. Thus an ability to earn very high prices from being able to react quickly to opportunities or provide capacity in an area that can become islanded and subject to sudden price surges/supply shortfalls tends to encourage desirable investment behavior. Requiring a window of several hours between bid and dispatch would frustrate this. By the same token, preventing firms from reacting quickly to cover their contracts in the event of an unexpected outage is likely to result in over-cautious holding back on capacity and higher costs. The abandonment of proposals to place restraints on re-bidding close to dispatch was influenced by observations that most of the very late changes to price in the market had been price reductions. These reflected decisions by marginal suppliers to bring plant on line in response to market opportunities that were unfolding (due to a demand surge, network constraint, etc.). Following the various investigations into market power, changes to the Code were made to recognize the possibility that a supplier could bid erratically and benefit from creating great instability to the market.16 Hence rules were tightened to require some explanation of Market supply in Victoria/South Australia 5000 4000 3000 2000 1000 s O

th

er

n er

he

ge

rn Sy n

ut So

N

R

G

y ow Sn

ng Ya

IP

y Lo

Ya (C llou LP rn )

0

Fig. 6.7. Source: NEMMCO. 16

This led to minor modifications to the guidelines on re-bidding volumes close to dispatch. See http://www.aer.gov.au/content/index.phtml/itemId/659216

190

Electricity Market Reform

re-bidding behavior. These changes were more of the nature of increased insurance against aberrant bidding behavior rather than affecting bidding patterns themselves. 6.3.4. Price outcomes in the wholesale market The resiling from major changes to the market rules was made more acceptable in view of the very low prices that had on average prevailed in the market. Compared to a notional wholesale price pre-market of around $A38/MWh in Victoria and $A40/MWh in NSW, competition has maintained average prices at low levels, below $A35 for most of the period. Higher prices prevailed early in the Queensland and South Australian regions as they were being bedded down. Table 6.3 illustrates the price developments in each of the regions. In addition, the forward price has tended to be stable, though edging up beyond the $A40 level by the end of the current decade, indicating a need, though not a pressing one, for new plant since at such prices new coal-based plant investment is profitable in Queensland, NSW, and Victoria. A synthetic forward price is published with the following indicative price for baseload energy (Fig. 6.8). Table 6.3. Average price ($A/MWh). Year

NSW

QLD

SA

SNOWY

1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 (2 months)

33.13 28.27 37.69 34.76 32.91 32.37 39.33 26.68

51.65 44.11 41.33 35.34 37.79 28.18 28.96 19.48

156.02 59.27 56.39 31.61 30.11 34.86 36.07 32.36

32.34 27.96 37.06 31.59 29.83 30.80 34.05 27.33

TAS

VIC

190.38 110.63

36.33 26.35 44.57 30.97 27.56 25.38 27.62 28.61

Source: NEMMCO.

Regional quarterly base futures prices 80

$/MWh

60 40

Queensland

New South Wales

Fig. 6.8. Forward baseload price. Source: NECA.

Victoria

Q4 09

Q3 09

Q2 09

Q1 09

Q4 08

Q3 08

Q2 08

Q1 08

Q4 07

Q3 07

Q2 07

Q1 07

Q4 06

Q3 06

Q2 06

0

Q1 06

20

South Australia

191

The Electricity Industry in Australia 6.4. Performance of the Reformed Electricity Market 6.4.1. Generation

The outcome of Australia’s reforms has been considerable improvements in productivity. In the case of power stations, increases in labor productivity since 1990 have ranged from a fivefold improvement in Victoria to a 50% improvement in Queensland, where it is generally acknowledged that power stations were operated more efficiently at the outset. Although there has been some new construction, the stock of generating capacity has not markedly changed over the period. Figure 6.9 illustrates the improved productivity.

Generator labor productivity (GWh/employee)

60 50 40

1990/1991 1996/1997 1999/2000 2000/2001 2002/2003

30 20 10 0

New South Wales

Victoria Queensland

South Australia

Tasmania

Western Australia (Western Power)

Power stations’ availability to run 100 95 90

1990/1991 1996/1997 1999/2000 2000/2001 2001/2002 2002/2003

85 80 75 70

New South Wales

Victoria

(b) Fig. 6.9. Source: ESAA.

Queensland

South Australia

Tasmania

Western Australia (Western Power)

192

Electricity Market Reform

In spite of these vast improvements in labor productivity, Australia’s power stations have also shown much greater reliability. As illustrated in Figure 6.9, the “Availability to Run” has improved in all of the States but especially in NSW and Victoria. This in itself has raised the de facto capacity of the industry. The comparisons indicate a relatively better performance on the part of those states that have privatized their businesses. In the privatized Victorian and South Australian systems, labor productivity has respectively increased fourfold and threefold. NSW with its government owned generation business and Queensland, under predominantly government ownership, have seen more modest increases – twofold in the case of NSW. In terms of labor productivity and plant availability, the privatized Victorian system now surpasses the performance of the state owned NSW system. This is in spite of Victoria having the disadvantage of relying predominantly on brown coal, which requires greater processing before being burned than black coal, which is the fuel source of the NSW system. Part, but by no means all of the relative Victorian improvement is due to greater use of contractors in the privatized firms. Moreover, to the extent that NSW uses fewer contractors this is likely to be a reflection of its shareholder’s preferences for union labor and would contribute, of itself, to lower levels of efficiency. There is also some indication of improved capital productivity in the privatized businesses. In particular, privatization brought a new lease of life to International Power’s 1600 MW Hazelwood brown coal generator in Victoria. The station had been previously scheduled to close in 2005 but has had its capacity increased and now likely to operate for another 20 years.

6.4.2. Distribution As with generation, distribution, which accounts for over 40% of final costs, has shown strong productivity improvements. Figure 6.10 shows that customers per employee in

Distribution businesses: customers per employee 900 800 700

1994/1995 2002/2003

600 500 400 300 200 100 0 New South Wales

Fig. 6.10. Source: ESAA.

Victoria

Queensland South Australia

Tasmania

Western Australia (Western Power)

The Electricity Industry in Australia

193

Victoria and NSW have almost doubled and all states, bar Queensland, have shown impressive gains. Again Victoria shows impressive labor productivity (the Western Australian system applies only to the interconnected, relatively urbanized system). In terms of outages the system performance has been mixed – in most areas outages are heavily influenced by occasional severe storms that do not occur with any degree of regularity. Figure 6.11 below indicates that outages have generally remained low and shown little trend, except in Victoria where they have been reduced and become comparable with outages in other states.

6.4.3. Consumer price outcomes Price level comparisons are distorted by the previous, and for households to a major degree on-going, regulation of maximum prices. Household retail price controls have suppressed prices in all states, though in recent years price controls have been eased or been allowed to rise in response to regulated increases permitted in line charges. This is especially the case in South Australia, where electricity costs are intrinsically higher than in other states and where the load profile is more skewed towards the summer peak than in other states. Household consumers’ prices in NSW and Queensland remain under relatively tight government control and are below those that would prevail in a commercial market. Over the past 7 years real prices for residential consumers have risen by as little as 9% in NSW and as much as 52% in South Australia. Those in Victoria have risen by 15%. As expected from a system with access to low-cost energy inputs, Australian prices remain among the lowest in the world. Business customers have been freed from price controls for several years. In the past, the business customers subsidized household customers but once the market became contestable this was no longer possible. As a result, real prices for business customers have fallen by over 23% in the three eastern states; they have risen about 2% in South Australia. Figure 6.12 below illustrates the trends.

Average outage duration (minutes off supply per customer) 350 300

1995/1996 1996/1997 1997/1998 1998/1999 1999/2000 2000/2001 2001/2002 2002/2003

250 200 150 100 50 0 New South Wales

Victoria

Fig. 6.11. Source: ESAA.

Queensland

South Australia

Tasmania

Western Australia (Western Power)

194

Electricity Market Reform

1996/1997 1996/1998 1996/1999 1999/2000 2000/2001 2001/2002 2002/2003 2003/2004 re NS si W de nt ia l N SW bu si lar ne g ss e re Vic si to de ria nt ia l Vi bu c l si arg ne e ss Q ue re e si ns de la nt nd ia l Q bu ld l si arg ne e ss S. re Au si st de ra nt lia ia l SA bu la si rg ne e ss

Cents/KWh

Real electricity prices (2001/2002 dollars) 20 18 16 14 12 10 8 6 4 2 0

Fig. 6.12. Source: ESAA.

6.5. Issues Concerning Public and Private Ownership The post 1990 period in Australia has seen dramatic improvements in the productivity of the ESI. Initially these strides were made under public ownership with reform of the overstaffed integrated utilities and their incorporation under company law. A further productivity improvement took place with the privatization of the Victorian industry from 1997. State owned businesses generally adopted many of the labor saving and production enhancing measures of their private sector counterparts. However the fact that the Australian industry remains mostly under public ownership carries several potentially damaging consequences. The first of these is the intrinsically greater incentives to save on costs that are present in privately owned businesses. Part of this may be due to government as the shareholder and appointer of the company boards, is more reluctant than a private company to shed labor; allied to this is close links that the Australian State Governments owners of the electricity businesses have with trade unions. This means they are reluctant to allow non-union labor and keen to ensure that union rights and privileges are maintained. In the main this will reduce the capacity of the managements to manage. The evidence available in Figures 6.9–6.11 indicates that the efficiency levels of the private sector businesses exceed those of the public sector firms. Public ownership also impedes firms from re-arranging their assets. All of the private sector businesses post-privatization have undergone several structural changes as parts of them have been spun off, other parts have been augmented and some have undertaken major new investments. Private sector businesses, in search of operational economies and improved shareholder value have re-arranged asset ownership so that activities are better grouped together. In the case of the distributor/retailers, this has led most of the private sector to have the two functions separated and housed in differently owned firms, whilst none of the public sector businesses have taken such steps. The public sector businesses need to approach their government shareholder for approval of any new capital investment. These decisions face the familiar issues of government decision taking. In New South Wales for example, the State Government has adopted an anticoal philosophy on greenhouse grounds (in June 2005 the “ecologically friendly” then State

The Electricity Industry in Australia

195

Premier even opened up the debate on nuclear generation, a debate that had been dormant due to opposition by green groups and because cheap coal offers a more economic baseload solution). By contrast to the NSW Government’s restraint on new investment, it is claimed that the Queensland State Government is accepting a lower than commercial rate of return in order to encourage the building of new power stations. It is certainly the case that most new capacity has been built in Queensland and about 60% of this has been government financed but Queensland also has the fastest growing load and has the cheapest coal. In this respect, data for Queensland (Table 6.3) shows spot prices were one third less than in NSW in the early part of 2005/2006. This may offer some corroboration of the over-build of generation in the State; however, the spot market data does not appear to be reflected in higher contract prices (Fig. 6.7). Whether or not Queensland new generation investment has been fully justifiable on profit grounds, public ownership adds a non-commercial dimension into the industry which diminishes the predictability of private firms’ competitive environment. Other things being equal, this brings additional business risk, excessive conservatism and higher prices/lower reliability.

6.6. Government Market Interventions 6.6.1. Mandatory insurance schemes Public ownership leaves governments with greater political vulnerability in the event of poor decision making. Hence decisions by both the NSW and the Queensland Governments to implement a form of mandatory insurance for the supplies to household retail customers. This, called the Electricity Tariff Equalization Fund (ETEF) in NSW, tends to blunten the market forces through reducing apparent retail risk by having the government assume much of it. One outcome is price suppression, especially for peaks, and a muted demand signal for new investment. ETEF operates by placing a ceiling and floor on wholesale prices as they impact on the household part of the aggregate load. When prices are high the generators receive only the stipulated price and reserves are accumulated in the fund; these are released when spot prices are below the floor set for the fund. As well as creating an insurance risk for the state, it is likely that such measures also impede the market for various financial instruments, the depth of which in NSW lags considerably behind Victoria. Although having similar features, the Queensland scheme, called the Long-term energy procurement or (LEP), has been designed with the aim of increasing liquidity and market depth by encouraging the incumbent retailers to actively seek contract cover for the franchise load. That is, the structure of the LEP is such that the incumbent retailers are exposed to energy price and volume risk and have incentives to manage their position through contracting, with compensation provided for efficient purchasing against a benchmark.

6.6.2. Retail price control Retail prices to business consumers have been deregulated in all states but all retain some controls over household prices. In the case of Victoria and South Australia, these controls are not considered to be significantly in excess of underlying market prices, and in both states there has been significant customer churn. This is also true to some degree of NSW and to a lesser extent in Queensland.

196

Electricity Market Reform

For electricity, retailers need to balance their demand and supplies and act as the agent of the final consumer. They pick up the tab when the price spirals out of control and this gives them a great incentive to ensure they measure their sales correctly and contract for supplies appropriately. They perform the same essential function in the energy market as elsewhere – they look to demand, seek to attract customers who they can profitably supply and package supply to meet their customers’ needs. The outcome is signals that drive efficient market activity. These signals include the prices that attract the right form of new supply (peak, off peak, etc.) They also develop prices that choke off or encourage increased demand. Of course, all this is made more difficult in the ESI. The absence of half hourly metering at the domestic level and the price cap are among the market realities that prevent this from operating with full effect. Even so, the residual retail price controls dampen the economic signals that businesses need to determine the optimum time to invest or undertake other strategic decisions. 6.6.3. Greenhouse gas abatement schemes Australian coal is inexpensive and located conveniently to major electricity loads. However, its economics change markedly in a Kyoto constrained electric power future, whether the measures in place to reduce carbon dioxide emissions are a form of cap and trade regulation, a carbon tax or even a more arbitrary set of regulations that have similar effects. Australian measures ostensibly aimed at reducing emissions of carbon dioxide appear in many forms and are inconsistent from one State to another. They impact most heavily on coal, Australia’s lowest cost energy source and include: ● ● ● ●

●

the federal government’s mandatory renewable energy target (MRET), the Queensland’s 13% gas target, the NSW’s Greenhouse Gas Abatement Certificate (NGAC) scheme, subsidies to wind and other exotic renewable sources offered through the Australian Greenhouse Office and state governments (the latter in the form of regulatory measures that reduce connection costs to wind generation), and schemes that mandate minimum energy savings on appliances. First applied to fridges and freezers and targeted at energy conservation, these regulatory requirements have been re-badged as greenhouse measures and extended to include houses as well as other appliances.

The MRET scheme’s focus is on renewable energy and requires retailers to acquire and annually surrender a progressively increased number of Renewable Energy Certificates (RECs). These essentially require usage of novel energy sources like wind, though some existing and expanded hydro capacity has managed to redefine itself to be eligible. By 2010, 9500 GWh (around 4.5% of demand) will be required under the MRET scheme. For the three schemes combined over 30,000 GWh is estimated to be covered amounting to over 13% of total demand.17 The Queensland scheme seeks to substitute gas for coal-based electricity inputs. The NSW scheme seeks to introduce a penalty on carbon dioxide graduated in line with the emissions per unit of energy of each electricity generation source. 17

Alan Moran. November (2004) Economic and Environmental Potential of Energy Efficiency Regulations: Submission to the Productivity Commission Inquiry into Energy Efficiency, at http://www.ipa.org.au/files/ Energy33.pdf

197

The Electricity Industry in Australia

The default penalty costs of the three regulatory measures provide a cap on the costs they are likely to entail. These costs entail a premium over the costs of conventional electricity to retailers. By 2010, when the schemes are at full maturity, the fall-back penalty rates for the Commonwealth, NSW and Queensland schemes respectively are $A40, $A14.3 and $A13.1/MWh. These rates provide the (maximum) subsidies to the non-carbon or low-carbon emitting fuels. The Commonwealth’s RECs during 2005 were trading 20% below the maximum rate. Table 6.4 below summarizes the more readily identified costs. A greenhouse trading regime has considerable support in Australia, with most state governments urging its national adoption. Figure 6.13 below estimates the costs of electricity with and without the sort of additional charges implicit if Australia adopted the EU cap and trade scheme and other schemes are left in place. Based on recent developments, we have relatively good information on the costs of conventional generation for the eastern seaboard of Australia. Nuclear costs have been rigorously Table 6.4. 2010 Costs of greenhouse gas support measures.

$AM

MRET

NSW NGAC

Qld 13% gas

Commonwealth subsidies

State subsidies

380

222

68

124 (2006/2007)

32 (2004/2005)

Source: Budget documents.

Comparative costs of power $90 $80 $70 $60 Cost per MWh Cost per MWh post rights

$50 $40 $30 $20 $10

Si uc m le ar N uc Sc ul m lea ly od r ul SA ar IC he G liu T m

d

N

in

G en

W N

uc

N

le ar

at

ur a

n

lg as

co al

ld Q ow Br

al co

co

Bl

ck Bl a

ac k

al

N

SW

$-

Assuming CO2 rights trade at $41/tonne Fig. 6.13. Australian power generation costs with EU tradable rights prices. Sources: Australian costs are based on recent estimates. Nuclear power costs are based on University of Chicago, The Economic Future of Nuclear Power. Carbon emissions per gigajoule of energy is derived from http://www.greenhouse.gov.au/workbook/pubs/workbook.pdf

198

Electricity Market Reform

evaluated in a recent University of Chicago report that brought together three contemporary estimates of the costs of nuclear generation (excluding the disposal costs). The models compared are the Shanghai automotive industry corporation (SAIC) industry model, the Scully Capital financial model, an Energy Information Administration (EIA) model and GenSim, which is based on the EIA approach.18 Wind power costs are also relatively well known. The price cited below ($A75/MWh) represents the costs at prime sites. These costs exclude any additional transmission charges that may be required for the more remotely located facilities. They also do not take into account the need for conventional power back up which is necessary once wind, with its unpredictable and intermittent nature, becomes a significant component of the aggregate supply. At present, coal-based generation costs is estimated to vary from about $A32/MWh in Queensland,19 around $A40 in NSW and $A38 in Victoria. Gas is estimated at around $A45/MWh based on a cost of $A4/Gigajoule, a cost that may rise if greenhouse measures raise the demand for gas. Nuclear costs exclude waste storage, the estimated costs of which vary. One estimate by the Uranium Information Centre puts the decommissioning costs as adding 5% to the price and waste disposal a further 10%.20 Applying the EU price of euros 24/tonne of carbon dioxide, cost to current power sources would result in nuclear power based on some cost estimates becoming marginally cheaper than coal and natural gas generation. Government greenhouse mitigation policies and laws have also been used by advocacy groups (themselves often government financed) to tie up new proposals in the courts. The highest profile case, which is discussed below, has been the NSW State Government using greenhouse policies to maneuver itself out of a (high-priced) contract for a new generator designed to use waste coal. Other cases have included the prevarication of the Victorian Government in granting approval for a major power station to have its life extended. Snowy Hydro illustrates an outcome of subsidies that cannot or are not fully defined to meet their stated goals. Snowy has been given an annual baseline level of generation above which it earns RECs that have a default value of $A40/MWh. Because it does not receive penalties for underperforming, it operates its system to generate strongly in 1 year and build up its reserves for the next year. This allowed Snowy to earn about $A67 million in essentially phantom RECs in 2003. Not only did this not contribute in net terms to the government policy of reducing greenhouse gas emissions, it actually increased emission levels. This is because 20% of Snowy’s RECs may have been created through pumping water uphill for reuse. Though the energy used in pumping is netted out from the REC creation, shifting water from an underperforming year to a hard generating year creates a credit. In pumping water uphill, Snowy uses almost twice as much (coal-derived) energy as it produces in subsequent generation. 6.6.4. Sovereign risk One further facet that is not considered in the comparative cost data is the sovereign risk involved in building fossil fuel power stations, especially involving coal. The activities of 18

See http://www.anl.gov/Special_Reports/NuclEconAug04.pdf Some private generators suggest this price, which is cited as being available from new government generators, is sub-economic price because it incorporates a de facto subsidy as a result of the Queensland government accepting lower levels of return than the private sector would require. 20 Source Uranium Information Centre http://www.uic.com.au/nip08.htm 19

The Electricity Industry in Australia

199

the NSW Government in using environmental pretexts to renege on coal-based contracts and the additional costs the Victorian Governments have required of the Hazelwood Power station upgrade proposals are likely to require a risk premium for coal powered electricity. The NSW Government has created obstacles and uncertainties in the way of new electricity generation. These make it inevitable that investors will require a risk premium before committing funds, thus increasing the price of electricity in the state and the risk of shortages. One example of this was the government’s treatment of a private sector investment undertaken by the US firm National Power. This stemmed from a commitment – in the event an unwise commitment – by energy Australia, the biggest retailer in the NSW for two power stations, Redbank 1 and 2. Soon after the deal was struck, the price in the market halved and remains 30% below the Redbank contract price. Some estimates put the contract loss at $A750 million. To renege on the deal for the second power station, the NSW Government set up an inquiry into it. Various Government funded green groups offered opposition to the project on grounds of its greenhouse gas emissions and the Government refused its development approval, thus avoiding an onerous contract. In fact opposition to the development on environmental grounds is ironical since the project uses waste coal which could otherwise pollute the Hunter River. Indeed, in 2001 Redbank 1 won the Institution of Engineers Award for Environmental Excellence. The Government’s performance on this matter must add to the risk premium required of private sector developers of power stations and will probably require some enforceable undertakings before any private funding is extended to coal-based generation in the state. In addition, private sector investors, especially in coal-based generation, would hardly be re-assured by the statements the NSW Government makes in a Green Paper issued in January 2005.21 This expresses considerable hostility to new coal power, canvasses greenhouse taxes and a policy (p. 24) by mid century ranging from stabilization of emission at current levels to reducing them by 40%. In terms of a business-as-usual growth in energy demand at 2%/annum, this range of outcomes would amount to a highly ambitious reduction of between 60% and 75% in emission levels. In claiming in concert with this, that the Government will let the market decide which technologies should be developed, the Green Paper is giving expression to grand sounding laissez-faire principles that clothe a highly intrusive policy approach. The Green Paper, due to proceed to a White Paper stage by June 2005, had not so progressed by October 2005. In the interim, the State Premier, Mr. Carr who was highly supportive of greenhouse gas reducing initiatives, unexpectedly retired.

6.7. Outcome of Australian Electricity Market Arrangements 6.7.1. Summary of market distortions Aside from sovereign risk associated with government de facto expropriations as occurred in the NSW Redbank case, distortions that could lead to serious market damage include: ●

21

The NSW mandatory insurance system or ETEF, provides a weaker incentive for retailers to ensure that they are forecasting market demand accurately. ETEF means the

NSW Government, Energy Directions Green Paper. http://www.deus.nsw.gov.au/new/NSW% 20Energy%20Directions%20Statement%20-%20702 KB.pdf

200

●

●

●

●

●

●

Electricity Market Reform

government has eliminated the risks to retailers of failing to forecast the household load accurately. This may bring mistakes caused by unexpected demand shifts. Retail price caps being kept below market levels. This is an area where governments in Victoria and SA have managed to control their propensities to intervene and are allowing prices to shift market levels. NSW however retains very low allowable retail margins, which seriously restrict competition. The risk that regulators will offer inadequate incentives for expansions and optimal maintenance. Major price reductions have been insisted upon by several state regulators. Regulated businesses are always likely to profess dissatisfaction at such outcomes but a risk remains that price cuts can deter investment. Such requirements may have been a feature of the fragility of distribution networks, especially in Queensland where the regulator demanded a 17% cost saving of the largest distributor, Energex, which was heavily criticized following power outages in 2005. Interventions favoring subsidized and uneconomic generation can suppress demand, which means reduced new investment especially in the sort of energy intensive industries that Australia is well placed to win. The various schemes like MRET and NGAC add costs to industry and in the case of NSW, mean that some 23% of electricity is now slated to be subsidized; this probably rules the state out of consideration for major new energy intensive industry. Less draconian measures are in place in other states – Queensland has its 13% gas requirement and Victoria was less than firm in controlling its state financed green groups who campaigned to prevent the Hazelwood expansion; nor has the state made a wise choice in its appointment of a relatively activist Presiding Judge to the Land and Environment Court whose legal interpretations prolonged the case and added expenses. Some private sector generation businesses claim that the new capacity building by the Queensland government is not based on commercial principles but are being subsidized indirectly by a government intent on using its cheap coal as an industry development tool. Though subsidized plant adds to capacity in the first instance, each new tranche of it considerably reduces the incentive of commercial parties to seek out opportunities to build plant in line with market requirements. Subsidized plant puts us on the slippery slide to total government ownership or control of the industry. Much the same risk has in the past been offered by subsidized transmission. If a power station is stranded by low-cost power being brought in from elsewhere it suffers lower than expected returns. If this is due to it being stranded as a result of government regulations that effectively subsidize costs, damage is done to the market’s automatic ability to supply demand. Finally, there remains the risk of other interventions. Australia has not moved to create trading barriers in response to fears about market power being exercised in re-bidding. Such measures would create major uncertainties by constraining generators’ abilities to respond to sudden emergency issues, would have gummed up the bidding system and created costs and uncertainties.

6.7.2. Outcomes in terms of new capacity and prices Notwithstanding the adverse effects of government intervention in Australia, a successful outcome has been observed. Prices are lower than expected, reliability has been high, and although the most market exposed sector, generation, has seen very low returns, new investment has been forthcoming. And the investment that has been made broadly corresponds

201

The Electricity Industry in Australia

to that which most experts expected: increased peaking capacity in Victoria and South Australia and more baseload to meet the faster growing Queensland demand. This demonstrates a resilience in markets. As long as the various participants in the market are free to contract with each other and as long as there is no significant monopoly over supply, interventions may not seriously distort the market and lead to its failure. Aside from wind power, which is totally dependent on subsidies, significant new power facilities built over the past 5 years are as shown in Table 6.5. Figure 6.14 illustrates how demand and supply have been fairly well synchronized over the past 6 or 7 years. The message is that the real dangers to the supply industry in both gas and electricity in Australia are those stemming not from too little government but from too much. The

Table 6.5. New capacity 2000–2005.

Redbank Bairnsdale ValleyPower Somerton Laverton Loy Yang Oakey Millmerran Swanbank E Tarong N Kogan creek Hallett Pelican point Ladbroke Quarantine

State

Capacity (MW)

Type

Ownership

NSW Vic Vic Vic Vic Vic Qld Qld Qld Qld Qld SA SA SA SA

150 92 300 160 312 236 282 852 360 450 750 220 320 80 100

Coal Gas Gas Gas Gas Coal Gas Coal Gas Coal Coal Gas Gas Gas Gas

Private Private Private Private Government Private Private Government/Private Government Government Government Private Private Private Private

Source: ESAA.

NEM generation and peak demand 45000 40000 35000 30000 NEM generation

25000 20000

Summer peak

15000 10000 5000 0 1998

1999

Fig. 6.14. Source: NEMMCO.

2000

2001

2002

2003

2004

202

Electricity Market Reform

industry has expanded and maintained low costs in the 6 years it has been operating. It is however fragile and government actions could seriously harm investor confidence and lead to ever increasing interventions to ensure investment keeps pace with demand. Such measures would gradually erode the massive lifts in productivity that has been observed over the past decade or so.

6.8. Glossary of Acronyms ACCC AEMC AER ESAA ESC ESI ETEF LEP MRET NCP NECA NEM NEMMCO NGAC ORR PC QEC REC SECV

Australian Competition and Consumer Commission Australian Energy Market Commission Australian Energy Regulator Electricity Supply Association of Australia (Victorian) Essential Services Commission Electricity Supply Industry Electricity tariff equalization fund (Queensland) Long-term energy procurement Mandatory renewable energy target National competition policy National electricity code administrator National electricity market National electricity market management company (NSW) Greenhouse Gas Abatement Certificate (Victorian) Office of the Regulator General Productivity Commission Queensland Electricity Commission Renewable Energy Certificate State Electricity Commission of Victoria

Chapter 7 Restructuring the New Zealand Electricity Sector 1984–2005 GEOFF BERTRAM School of Economics and Finance, Victoria University of Wellington, Wellington, New Zealand

Summary This chapter describes New Zealand’s failure over two decades of reform to establish a viable industry self-governance framework, and the parallel failure to achieve restraint on monopoly profits by means of light-handed regulation. Starting from a classic publicly owned monopoly of generation, transmission, distribution, and retailing, New Zealand corporatized all levels of the supply chain, separated lines businesses from generation and retail, removed retail franchises, and broke up the monopoly generator into five companies, two of them privately owned. These measures were insufficient to achieve competitive outcomes in the absence of hands-on regulation. Generators integrated vertically by takeover of retailers, and the resulting retail oligopoly erected an effective barrier to entry by withholding affiliated generators’ capacity from the very thin market for hedge contacts. Grid pricing and contract provisions foreclosed demand-side innovation and distributed generation. Distribution lines businesses ramped up mark-ups from 30% to 70% without any regulatory restraint, and were allowed to revalue their assets to underwrite the new high margins. Faced with failure of the original design, the Government in 2003 established a new industry regulator and invested in new state-owned thermal generation to plug the country’s yawning gap in reserve capacity.

7.1. Background New Zealand is a country of 4-million people spread across an area the size of Italy or the UK. From south to north the country is over 2000 kilometers in length, with the two main islands separated by the 30-kilometer-wide Cook Strait. The largest city (and major electricity load center) is Auckland, with 1.3 million inhabitants. Electrification began in the late 19th century, when local authorities and private entrepreneurs constructed small generation facilities to serve local markets.1 Following the First World War the Government embarked on the construction of a set of large state-owned 1

A detailed history of the New Zealand electricity industry is Martin (1998). See also Rennie (1988), Jackson (1988, 1990).

203

204

Electricity Market Reform

hydroelectric plants on major rivers, linked by a transmission grid from which power was taken off by local-government distribution and retail companies (Electrical Supply Authorities, ESAs), each with a territorial monopoly franchise. ESAs supplied a bundled service, comprising low-voltage distribution networks, the retailing of electricity to final customers, and supply and servicing of household electrical appliances. In the 1950s, when major new investments in generation plant struggled to keep pace with demand growth and blackouts were a common occurrence, most households were placed on ripple control to switch off water heaters at times of peak demand.2 The state-owned generation and transmission system built up from the 1920s displaced most locally owned generating plant, and standardized the countrywide retail supply voltage at 220/240 volts at a frequency of 50 MHz (matching the UK settings). For the next halfcentury, electricity generation and transmission remained a state-owned monopoly, while distribution and retail remained franchised, publicly owned, local monopolies. Regulation in this setting was redundant, since both central and local government were democratically accountable, and operated the electricity supply system with social, rather than commercial, goals. Prices were set to achieve break even, in cash flow terms, over the long run. Financial disclosure, in terms of the cash flow model used for much of the public sector, was comprehensive, with detailed accounts for all levels of the system published annually.3 Asset values were recorded in historic-cost terms without adjustment for inflation, and were also lowered by the common practice of expensing day-to-day small-scale acquisition of capital equipment. Initially the two main islands had separate electricity grids, but there was an obvious mismatch between the abundant hydro resources of the South Island and the concentration of load in the North Island, particularly in Auckland. In 1965 a high-voltage direct current (HVDC) cable across Cook Strait connected the two systems together, allowing power from large hydroelectric developments in the South Island – particularly Benmore (540 MW) and Roxburgh (320 MW), on the Waitaki and Clutha Rivers, respectively – to be sent north. Thereafter the entire national generation and transmission system developed as a single integrated whole. The North Island accounts for around two-thirds of national demand but only one-third of generating capacity; the South Island has two-thirds of generation capacity but only one-third of demand.4 New Zealand’s annual electricity consumption is currently around 36,000 GWh, supplied from a system with 8500 MW of installed capacity. The 50% capacity utilization ratio reflects

2

Ironically, this almost universal penetration of simple demand-management technology in the period of public-sector monopoly has been allowed to slide away in the era of “market reforms” since 1987, as large commercially oriented firms on the supply side have welcomed demand-driven price spikes which they could take directly to their bottom lines. 3 The Minister in charge of the New Zealand Electricity Department (NZED) tabled a full annual report in Parliament each year. All ESA financial and operational data was published annually from the early 1960s under the cumbersome title Annual Statistics in Relation to Electric Power Development and Operation for the Year Ended 31 March. The latter publication rapidly reduced its coverage in the early 1990s and was discontinued in 1994. Its successor, the company-by-company regulatory information disclosure from 1994 on, was both less informative and inconsistent from company to company, which means that public monitoring of performance has been more difficult after the reforms than before. 4 The mismatch between the two islands would have been greater still had it not been for the establishment in the 1960s of the large Comalco aluminium smelter at Bluff in the far south, which by itself comprises about 17% of national demand and provides the principal market for the Manapouri hydro scheme, the country’s largest with capacity of 710 MW (upgraded from 585 MW in 2002).

205

Restructuring of the New Zealand Electricity sector 1984–2005 Table 7.1. Trends 1965–2004.

1965 1970 1975 1980 1985 1990 1995 2000 2004

Total installed generating capacity (MW) 2336 3683 4784 5860 6988 7067 7910 8845 8515

Peak load (MW)

Total consumption (GWh)

2048 2690 3391 3677 4642 5122 5240 5830 6090

8189 11,069 16,272 19,040 23,994 27,309 29,925 32,735 35,795

Total sales revenue ($m)

Average final price (c/kWh)

Real average price, c/kWh at March 2004 prices

90.0 143.3 196.4 681.5 1190.4 2144.2 2490.2 2888.2 4014.5

1.10 1.29 1.21 3.58 4.96 7.85 8.32 8.82 11.22

14.56 13.79 8.41 12.58 9.58 10.64 10.23 10.36 11.85

Sources: Installed capacity from Annual Electricity Statistics and Energy Data File for years shown. Consumption, revenue, and prices from Energy Data File January 2005, p. 126 Table G.12, p. 134 Table I.1, and p. 135 Table I.2. Real average price 1965–1975 derived using CPI.

the fact that two-thirds of supply comes from hydro generators which are designed to run at a low load factor, combined with the existence at the margin of some high-cost thermal generating capacity which is operated for only part of the year. System-wide capacity utilization has risen steadily over recent decades, reaching 40% in the mid-1980s and approaching 50% in the mid-2000s. Table 7.1 sets out key statistics of capacity, consumption, revenue, and final price from 1965 to 2004. This period includes the last two decades of the old system, the “reform” years from 1986 to 1998, and recent experience with the restructured system. Figure 7.1 shows installed capacity and peak load since 1964. Capacity growth has proceeded in a stop-start fashion, attributable partly to the lumpiness of generation projects, partly to swings in policy, and partly to commercial decisions since corporatization. In the mid-1960s the momentum of the hydro construction program was at last outstripping demand growth after a decade of stress in the 1950s. System peak load in the mid-1960s was around 90% of installed capacity, but with hydro capacity expanding 8.5% per year until the mid-1970s, the ratio was brought down to below 70% by the late 1970s, and has remained around that level for the subsequent three decades. Peak load growth, which caused concern among power planners in the 1970s and 1980s, slowed down from the late 1980s; the central problem since 1990 has been maintaining supply in dry years. Figure 7.1 shows also a slackening in the pace of new construction following deregulation in the early 1990s, and the impact of the periodic decommissioning by the new owners of commercially unattractive dry-year-reserve thermal plant, which has left the system increasingly exposed to climatic fluctuations. The map of the main high-tension transmission grid in Figure 7.25 shows the location of the two main bottlenecks in the transmission system: the HVDC link from Benmore to Haywards, and the central North Island between Haywards and Otahuhu. For the purpose of understanding the basic economics of the network, the nodal spot prices at these three key measurement points suffice to put a price on the two key transmission constraints, which cause market segmentation into three main regions at times of stress (Videbeck, 2004). 5

For a detailed map of the entire grid showing all nodes, see http://www.transpower.co.nz/?id⫽4631

206

Electricity Market Reform

10,000 9000

Meremere (coal) closed

Marsden A (oil) closed

Otahuhu B and Taranaki CC (gas) commissioned

Clyde (hydro) commissioned

Stratford, Otahuhu A, and Whirinaki (gas/diesel) decommissioned Manapouri (hydro) expanded

8000 7000

MW

6000 5000 4000 3000 Capacity Peak load

2000 1000 1960

1965 1970

1975 1980

1985

1990 1995

2000 2005 2010

Fig. 7.1. Generating capacity and peak load, 1964–2004. Source: Data compiled from Annual Statistics in Relation to Electric Power Operation in New Zealand 1965–1993, and from Energy Data File for years 1995–2004.

Auckland Central North Island constraint

Otahuhu Huntly

New Plymouth Cook Strait constraint

Bunnythorpe Haywards substation

Major generation and substations Main load centres HVDC link

Wellington

Manapouri

Christchurch

Main high-voltage AC transmission lines Key constraints

Benmore Dunedin Comalco smelter

Fig. 7.2. Major transmission lines, showing location of the two main constraints.

Restructuring of the New Zealand Electricity sector 1984–2005

207

7.2. Supply/Demand Balance Three key features of the electricity supply industry (ESI) in New Zealand have to be borne in mind when considering options for restructuring: ● ●

●

Most generation (60–70%) is from renewable sources (hydro and geothermal). The hydro lakes are located mostly in steeply sloping river valleys and provide storage capacity for only a few weeks, which means that unusually dry climatic conditions quickly translate into reduced supply. Similarly, unusually high inflows of water must be utilized within quite a short time horizon, or else be spilled to waste. New Zealand is a stand-alone closed market, with no means of importing or exporting electricity. A supply shortage, therefore, results directly in demand rationing and/or price spikes, while excess potential supply can be neither stored beyond a short period, nor sold in external markets.

Prior to the restructuring, which began in the mid-1980s, New Zealand’s generation plants were operated on the basis of control procedures that equated the shadow value of stored water to the short-run marginal cost of thermal generation. So long as river flows were adequate, hydro plant could be operated as baseload, with thermal peaking plant utilized in periods when demand exceeded the supply available from optimal utilization of water. The usual roles of hydro and thermal generation were thus reversed. However, hydro also performed (and still performs) the very short-run task of frequency control, via the Maraetai II generating station on the Waikato River.6 Until the early 1990s the state-owned monopoly generator and grid operator, the New Zealand Electricity Division (NZED, later the Electricity Corporation of New Zealand, ECNZ), carried out this optimization exercise internally, and scheduled its various generation facilities to optimize the utilization of water by attaching a shadow value to hydro generation to reflect both foregone opportunities to utilize water in later periods, and planners’ judgments regarding future hydrological conditions. If lakes were full and high inflows were expected, hydro plant would be operated at capacity. If lake levels were low and a dry year was anticipated, water would be held back and more thermal plant brought online to fill the resulting gap in supply. Unchallenged control of a balanced portfolio of generation options enabled NZED to reap economies of scope as well as scale, because of its ability to internalize spillover externalities amongst various generation technologies. In particular, the explicit balancing of hydro and thermal generation options to maximize year-round operating efficiency of the system as a whole was the key to the ability of NZED to provide a very high level of security, and quality, of supply across the entire country, even in the face of climatic variability (mainly uncertainty about rainfall and, hence, river flows). NZED’s explicitly forward-looking scheduling and planning procedures took advantage of this heterogeneity of its generation assets to supply wholesale power at an average-cost price (the bulk supply tariff, BST), with operating surpluses from hydro generation used to cross-subsidize the high-operating-cost thermal firming plant. From 1957 on, the BST

6

The two generating stations attached to the Maraetai dam have a total capacity of 360 MW, well in excess of the capacity needed to utilize run-of-the-river flow. The second station (excess capacity) installed in 1971 was designed to provide frequency control for the national grid, and has metering and control equipment to detect and offset load fluctuations. See http://www.mightyriverpower.co.nz/ Generation/AboutUs/HydroStations/Maraetai/Default.aspx

208

Electricity Market Reform

included a levy on consumers to fund new investment in generation and transmission as well as covering operating costs of the system. This cash-in-advance approach meant that whenever a major new round of investment was undertaken, the BST would be raised to provide the necessary funds in advance. Consumers were immediately conscious of the resulting rise in retail charges, which meant that electricity investment was always politically sensitive. The managers of the system were motivated both by the quest for engineering efficiency, and by this political sensitivity, since NZED and its controlling Minister would carry the political blame for any supply outages. There were strong incentives to invest ahead of demand,7 keeping a substantial safety margin in both generation and transmission; but there was a countervailing possibility of political backlash if excessive investment programs drove up the BST, and hence the price to consumers, unduly. In the late 1970s and early 1980s the system’s planners maintained a wide margin of excess capacity and embarked on a major round of large hydro construction, which exposed NZED to criticism that it was over-investing relative to a socially optimal benchmark. Such criticism was particularly acute in the mid-1980s as it became apparent that the momentum of the ongoing hydro construction program had carried NZED into a series of large hydro projects (Tongariro, Rangipo, and Clyde) whose unit costs were orders of magnitude higher than the BST. As it became generally accepted that the long-run marginal cost (LRMC) of new generation had risen sharply relative to the average cost of supply, a noisy debate ensued between advocates of immediate increases in the wholesale electricity price to signal future costs, and supporters of continuing with the long-established average-cost pricing approach. This pricing debate is discussed further below.

7.3. Restructuring the Sector 7.3.1. First steps There was a sea change in New Zealand economic policy in the mid-1980s, as neoliberal economic doctrines (largely copied from the UK) were adopted by key ministers in the Labor Government elected in mid-1984, resulting in radical changes to all state-owned operations, including electricity. Initially the aim was to ensure that state-owned monopolies increased their profitability by raising their prices to contribute to reducing the government’s budget deficit (Ministry of Energy Financial Objectives and Pricing Review Team, 1984). A second goal, initially also motivated by revenue maximization rather than structural reform, was to raise the economic efficiency of state-owned operations by converting them into profit-oriented commercial corporate organizations. Linked to this was a desire to curb what were perceived by the New Zealand Treasury at the time as excessive investments in new capacity, which officials regarded as a drain on scarce government resources. In 1986 the Labor Government announced its decision to reform publicly owned trading activities, including the generation and transmission sectors of the electricity industry,8 and a State-Owned Enterprises Act was passed to govern the process of corporatization.

7

As Chapter 1 notes, many countries have had difficulty with investment incentives in the new restructured environment. 8 For a detailed official history of the reforms summarized here, see Chronology of the New Zealand Electricity Reform, at http://www.med.govt.nz/ers/electric/chronology/index.html

Restructuring of the New Zealand Electricity sector 1984–2005

209

In April 1987, the NZED was converted into the ECNZ and a private-sector entrepreneur was recruited to head the new board. The following year the operation of the transmission grid was transferred to a new ECNZ subsidiary, Transpower Ltd, as a first step toward separation of generation from transmission. The expectation of key policy-makers was that the generation assets of ECNZ would in due course be privatized, while the grid would be separated off under a governance arrangement that would restrain its exercise of market power. In December 1987 the Government set up an Electricity Task Force to advise on the new industry structure and regulatory requirements. The Task Force reported in September 1989, with three key recommendations: establishment of a competitive generation market, separation of the Transpower grid from the ECNZ, and introduction of competition at retail level. Box 7.1 lists the detailed recommendations.

Box 7.1 1989 Task Force Recommendations Generation ●

●

●

Generation entry barriers should be minimized and a regulatory rule against price discrimination by ECNZ be explored. Large-scale break up of the generation system is not favored but it is recommended that further study of the costs and benefits of spinning off one or two competitive generating companies be undertaken. Subject to satisfaction on competitive pressures in the generating sector, ECNZ should be privatized.

Transmission ● ● ●

The ownership of transmission assets should be separated from the generator. Distributors and generators should form a club to own the transmission grid. The regulatory framework for transmission performance monitoring should provide recourse to and reliance on intervention provisions in the Commerce Act 1986.

Distribution ●

●

●

●

Removal of franchise areas for the supply authority monopoly distribution and retailing of electricity, this to be combined with the removal of the obligation to supply. Tariffs to consumers should show transmission and distribution costs separately from energy costs. Supply authorities should be corporatized and subsequently privatized for listing on the share market. No regulation of retail energy prices, and regulation of distribution line charges should be “light handed”.

Source: Report of the Electricity Task Force, 1989.

210

Electricity Market Reform 100%

Share of generation Generation

ECNZ Transmission

Distribution ESAs Retailing

Consumers

Residential, commercial, agricultural, small and medium industrial consumers

Large directsupply industrial consumers

Fig. 7.3. Electricity industry structure 1990.

The last of these recommendations, namely no price regulation, and adoption of a light-handed approach to regulation in general, was wholeheartedly adopted by the Government. New Zealand’s early decision not to set up an industry regulator, and to rely solely on general competition law (the Commerce Act 1986) to protect the competitive process and the interests of consumers, distinguished subsequent experience sharply from that in Australia where a specialist regulator was established. Until recently, Germany (see Chapter 8) was the only other OECD country to embark on electricity restructuring without a specialist regulator. Both New Zealand and Germany have now established such regulators.

7.3.2. Initial structure Prior to restructuring, there were two tiers in the electricity sector: the NZED, a government department controlling all large generation and the high-tension transmission grid; and a large number of ESAs running low-voltage distribution networks bundled with retail energy sales and appliance sales and service. A limited number of large industrial customers took supply direct from the grid; all other final purchasers were customers of local franchise-monopoly ESAs. NZED delivered wholesale electricity (bundled generation and transmission) to distributors at a bundled price (the BST). The pre-reform structure is shown in Figure 7.3. Distributors set prices to recover their costs, with price discrimination in favor of domestic consumers (low priced) relative to commercial customers (high priced) and industrial customers (in between). This price discrimination may have been Ramsey efficient,9 but 9

Residential electricity demand is probably more elastic than commercial, because of households’ ability to switch to alternative fuels such as gas, coal, and wood.

Restructuring of the New Zealand Electricity sector 1984–2005

211

was portrayed by reformers as being due solely to politically motivated cross-subsidies in favor of residential users (Jackson, 1990).10 Although it was a dominant monopoly, the NZED prior to the mid-1980s exercised its market power only in pursuit of a politically set target of covering costs and collecting a margin sufficient to fund new investment projects. Similarly, ESAs had secure monopoly franchises in their territories but their boards were accountable to consumers via regular elections, which had the effect of maintaining continual pressure on management to maintain high standards of supply and to seek only small profit margins. The restructuring timetable over the two decades from 1984 is summarized in Table 7.2. 7.3.3. Generation and transmission restructuring Change began with corporatization of the NZED in 1987 to form ECNZ. In 1994 generation was fully separated from transmission, leaving ECNZ with generation while the transmission grid company Transpower became an independent state-owned enterprise charged with operating the grid and scheduling the dispatch of generators. Thereafter, in a series of steps from 1996 to 1999, the larger ECNZ generation assets were split up among four successor companies: Contact Energy, Meridian Energy, Mighty River Power, and Genesis; while the smaller ECNZ stations (plus a number of other generation plants formerly owned by supply authorities) were privatized by sale to Trustpower, Todd Energy, and two smaller operations owned by Natural Gas Corporation (NGC) and Tuaropaki Power. Contact Energy was privatized by a share float in March 1999; the other three large successor companies remain state owned.11 By 2004 these were the eight generator class members of the New Zealand Wholesale Electricity Market.12 The evolving market shares of the main generators, as measured by capacity, are shown in Table 7.3. Two generators, Contact and Meridian, between them now account for 57% of installed capacity, with the remaining 43% distributed among the other six players. An important consequence of the break-up of the ECNZ generation portfolio was that some complementarities among different types of generation in the formerly integrated system were lost. Of the successor companies, Genesis was heavy on thermal plant and light on hydro; Meridian and Mighty River initially had only hydro and wind generation, with no thermal;13 Trustpower’s portfolio of small plants comprises entirely hydro and wind. The only operator to inherit a diversified generation portfolio was Contact Energy, the first firm to be split off from ECNZ and privatized. Contact’s ownership of large North Island thermal (at New Plymouth, Otahuhu, and Stratford), large geothermal plant at Wairakei and Ohaaki,

10

A feature of reform rhetoric in the early 1990s was the alleged need to eliminate “cross-subsidies” by lowering commercial tariffs and raising domestic ones. No evidence of the relative demand elasticities of these groups was ever publicly advanced to demonstrate that the prevailing price relativities were not Ramsey efficient. The elimination of retail price differentials in the 1990s was driven more by commercial-sector political lobbying than by economic analysis. 11 There is no evidence to date that the state-owned companies have performed any differently from the private ones. 12 http://www.nzelectricity.co.nz/C2bMarket.htm 13 Mighty River subsequently took over a 125 MW gas cogeneration plant at Southdown, and was vested with ownership of the mothballed (never commissioned) Marsden B station, which it is now planning to convert to coal.

212

Electricity Market Reform

Table 7.2. Major milestones in the New Zealand reform process. Event

Date

Comments

Pricing review

1984

ECNZ established

1987

Electricity Task Force Partial grid separation

1987 1988

Task Force Report

1989

Ministry of Energy abolished ESA corporatization announced Transpower Establishment Board set up Transpower Establishment Board report

1989

Officials sought revenue gains from increasing electricity prices. Corporatization of the state-owned generation and transmission system. Task force set up to design restructuring program. Transpower set up as ECNZ subsidiary to be grid and system operator. Recommendations: privatize generation and distribution, separate the grid as a club, end distribution franchises, adopt light-handed regulation. Removed Government’s in-house specialist resource, hence lowered policy and analytical firepower available to Ministers. ESA boards converted to trustees, commercial directors appointed. To implement Task Force recommendations re grid restructuring.

1990 1990 1991

Adopted the novel optimized deprival value methodology to value assets at separation from ECNZ; stuck with club ownership proposal. Energy Companies Act 1992 Distribution companies (ESAs) to be corporatized. Parliamentary Select 1992 Rejected ECNZ case for wholesale price increases; recommended Committee report adoption of progressive (increasing block) pricing of power. on pricing Echoed by private sector “Hydro New Zealand” proposal (Terry et al., 1992). Winter supply crisis 1992 May–July drought caused blackouts; ECNZ water allocation criticized. Committee of Inquiry 1992 Investigated the winter crisis, recommended greater security margins. WEMS report 1992 Private-sector proposals for generation restructuring and pricing. WEMDG set up 1993 To advance WEMS agenda for competitive pricing and by Government wholesale market. Electricity Market Co 1993 New company established to manage and monitor a wholesale market. Retail franchises 1993–1994 First small consumers, then large consumers open to retail removed competition. Full grid separation 1994 Transpower becomes a State Owned Enterprise SOE; club proposal abandoned. Disclosure regulations 1994 Information disclosure becomes mandatory for all lines businesses; accounting separation of retail and lines activities. WEMDG report 1994 Recommended competitive pool and spot market, separate grid, long-term tradable wholesale contracts, restrictions on ECNZ market power. Generation split up 1995 ECNZ to be split in two, small hydro to be privatized. Contact Energy 1996 Separate SOE generator set up with 25% of ECNZ’s generation assets. MARIA established 1996 Industry arrangements to resolve competitive reconciliation issues at retail level. Wholesale market 1996 Pool, spot price, wholesale market come into being.

Restructuring of the New Zealand Electricity sector 1984–2005

213

Table 7.2. (Continued) Event

Date

Comments

Auckland CBD event

1997

Line/energy separation ECNZ split announced

1998

Contact Energy privatization announced ECNZ split carried through MACQS agreement Ministerial Inquiry

1998

Distribution company’s line into central Auckland city fails. Recurrent blackouts, emergency new line built by Transpower. Deferred maintenance probably a contributory factor to the breakdown. All ESAs forced to divest either their retail or their lines businesses. ECNZ to be split into three state-owned generators at April 1999. Shares floated in March 1999; cornerstone 40% to Edison Mission. Now four major generators plus privatized small hydro.

1999 2000

Governance Committee

2000

Electricity Industry Bill

2001

Winter supply crisis

2001

On Energy bankruptcy

2001

Hydro spill reporting Market bids and offers disclosure Light-handed regulation fails Another dry-year looms

2002 2002

Targeted regulation

2003

Electricity Commission

2003

New regulatory framework for grid investment and pricing New market arrangements Whirinaki opens

2004

1998

1999

2002 2003

2003 2004

Electricity Governance Rules

2004

Core grid defined

2005

Industry self-governing arrangement for grid security. Reported on regulatory issues; gave lines businesses a clean bill of health. Electricity Governance Establishment Project to create a unified self-governing framework. Made provision for direct regulation of lines businesses and Government imposition of governance arrangements if industry failed to self regulate. July–September shortage due to low lake levels. Blackouts averted by voluntary savings achieved by publicity campaign. Last independent retailer driven out, all retailers now vertically integrated with generators. Hydro generators must report any spillage to waste. Full detailed information to be published with a 4-week delay. Commerce Commission retrospectively legitimizes lines businesses’ asset revaluations. March–June predictions of a dry winter, and Contact’s withdrawal of some thermal capacity, led to major spot-price spike in April. Commerce Commission moves toward regulation of lines businesses. Industry regulator set up to organize governance, oversee supply security, build and contract for reserve thermal, regulate prices. Electricity Commission to coordinate new investments in grid and generation. Electricity Commission takes over the running of the sector under new rules and regulations. New state-owned reserve generator to underpin security of supply. New governance framework decreed by Electricity Commission after industry participants fail to reach agreement. Commission identifies a subset of grid assets which must meet very high reliability standards to avoid “cascade failure”.

214

Electricity Market Reform

Table 7.3. Generator shares of capacity, 1994–2004. 1994 Firm ECNZ Contact Genesis Mighty River Meridian Trustpower Others Total

1998

2004

Capacity (MW)

Percent of total

Capacity (MW)

Percent of total

7391

95.9

5361 2046

66.2 25.2

696 8103

8.6 100.0

317 7708

4.1 100.0

Capacity (MW)

Percent of total

2448 1541 1266 2539 452 474 8719

28.1 17.7 14.5 29.1 5.2 5.4 100

Sources: 1994 from Electricity Enterprise Statistics 1994, pp. 24–25. 1998 from ECNZ Annual Report 1997, p. 31; Contact Energy from Prospectus dated March 31, 1999, p. 21. 2004 from Ministry of Economic Development Energy Data File January 2005, pp. 116–119.

and two of the largest South Island dams on the Clutha River, has endowed it with greater ability than its competitors to schedule its generating plant strategically.14 The wholesale electricity spot market, set up in 1996 and run by the Marketplace Company (M-Co), is based on the interaction of supply and demand.15 The final price is equal to the last offer price necessary to meet demand, in a single-price auction where all generators receive the same final price regardless of their bid prices. A constraint-adjusted spot price is then set for every half-hour at approximately 250 “nodes” on the national grid.16 In theory, each nodal price is optimized to achieve the lowest overall cost to the country as a whole, given the offers into the pool by generators.17 There has been a sharp contrast between the adoption of complex and sophisticated pricing mechanisms on the supply side of the wholesale market and the almost complete absence of scope for economic incentives to operate on the demand side. The system operator treats demand as completely price inelastic, and there is no mechanism by which either electricity saving by consumers or small-scale distributed generation can participate in the wholesale market from the demand side. In the dry-year crises of 1992 and 2001 the Government resorted to mass publicity campaigns urging voluntary savings by consumers, but at no stage have economic rewards been offered for conservation effort.18 The New Zealand electricity reforms have been notable for the absence of initiatives such as real-time retail pricing to reward conservation effort by consumers, and opportunities for small-scale distributed generators to enter the market.19 14

A detailed history of Contact Energy in New Zealand, from an avowedly critical point of view, is at http://www.converge.org.nz/watchdog/08/06.htm 15 NZEM Pricing, www.nzelectricity.co.nz 16 NZEM Pricing, www.nzelectricity.co.nz 17 NZEM Pricing, www.nzelectricity.co.nz 18 An exception may be the Comalco aluminim smelter, whose contract with Meridian Energy is confidential but is rumored to include a provision for interruptibility. 19 There is a strong contrast between New Zealand’s effective foreclosure of small distributed generation and Tasmania’s well-established policy of purchasing power from individual consumers who have installed photovoltaic equipment on their properties; see http://www.auroraenergy.com.au/askaurora/ solarpower.html#Anchor-You-33869

Restructuring of the New Zealand Electricity sector 1984–2005

215

The absence of initiatives toward providing consumers with time-of-use metering and pricing, or net-metering arrangements for consumers with small-scale generation of their own, has contributed to the inexorable growth of demand and stands in sharp contrast to the finetuned and complex system of pricing signals on the supply side. Perhaps most striking has been the adoption of a detailed nodal pricing system for the delivery of power off the grid. 7.3.4. Nodal pricing A feature of the New Zealand reforms has been wholehearted adoption of the concept of detailed nodal pricing (Hogan, 1992, 1999; Ring et al., 1993a, b), with the result that there are no fewer than 250 separate nodal prices posted across a grid with only 480 entry and exit points. Much of this detail seems redundant to effective functioning of the market, and on balance has probably impacted negatively on market efficiency.20 Until recently there have been only two important bottlenecks in the New Zealand grid (see Fig. 7.2): the inter-island HVDC link, and the central North Island. (In the near future the latter constraint will shift northward to the transmission lines between Huntly and Auckland, once a planned new large thermal generator at Huntly is commissioned.21) The three key nodes in the system are Benmore (at the southern end of the HVDC link), Haywards (at the northern end of the HVDC link), and Otahuhu, in Auckland (north of the mid-North Island bottleneck). Figure 7.4 shows that the spot prices at these three key nodes move quite closely together, although from time to time one or other of the two transmission constraints binds, causing regional prices to diverge. These divergences, however, are of second-order significance relative to the overall volatility of the wholesale spot price. Price divergences at the other 247 nodes are generally insignificant. From time to time, the three principal nodal prices become separated due to grid constraints. During October 2000, for example, when the mid-North Island constraint was tight, the Otahuhu nodal price was roughly double the Haywards price, while Haywards and Benmore tracked closely together. Similarly, in January 2003, the Haywards price of 3.58 cents (c)/kWh became 5.03 c/kWh at Otahuhu, a difference of 41% from south to north of the North Island.22 An example of the HVDC constraint binding occurred in January 2002 when the Benmore price of 1.61 c/kWh was nearly doubled to 2.98 c/kWh at Haywards.23 Again in December 2002, the Benmore price of 3.65 c/kWh became 4.94 c/kWh at Haywards, and 6.12 c/kWh at Otahuhu.24

20

It could be argued that the design and implementation of the detailed nodal pricing arrangement has been driven primarily by engineers and consultants for whom the issue has been both lucrative and technically interesting. 21 Inspection of Figure 7.2 shows that major generation at or north of Huntly will be downstream of the central North Island constraint and will thereby relieve it. However, expanded transmission capacity will then be required between the new generator and the Auckland market. The siting of the new transmission line is at present embroiled in a difficult resource consent process. 22 Figures for the examples of constraint pricing here are taken from http://www.nzelectricity.co.nz/ electricity_prices/finals2003/August2003ReferencePrices.xls 23 See NZEM, Wholesale Electricity Prices Report 19 February 2002, at http://www.electricity.co.nz/ C2dPricesMonth/020219.htm 24 The main grid constraints can also bind in the opposite direction, at times when water shortages in the South Island require electricity to move south rather than north. For example, in August 2001 (a crisis period in a dry year with South Island hydro operating well below capacity) the average Otahuhu spot price was 9.93 c/kWh, the Haywards price was 11.13 c/kWh, and the Benmore price was 12.73 c/kWh.

216

Electricity Market Reform

300 250

Dry-year winter shortages June– August 2001

April 2003 price spike

$/MWh

200 150 100 50

N

ov Ma em y b 19 D J er 1 98 ec u 9 em ne 98 be 19 r 1 99 J Ja uly 99 nu 2 9 Au ary 000 gu 20 Se M st 01 pt arc 20 em h 01 be 200 r 2 O Apr 20 ct il 02 ob 2 e 0 N M r 2 03 ov a 0 em y 03 b 20 D J er 04 ec un 20 em e 04 be 200 r2 5 00 5

0

Otahuhu Haywards Benmore

Fig. 7.4. Monthly average spot price at three main nodes, 1999–2004. Source: Data from http:// www.stat.auckland.ac.nz/⬃geoff/elecprices/

These examples, however, are not typical of the day-to-day functioning of the system. In a month of normal operation, with no significant constraints apart from line losses, and with power moving north on the HVDC link, the three main nodal prices converge quite closely. In May 2005, for example, the Benmore spot price averaged 6.89 c/kWh, the Haywards spot price was 7.03 c/kWh, and the Otahuhu spot price was 7.09 c/kWh, an overall differential from south to north of only 3%. 7.3.5. Distribution and retail restructuring The Energy Companies Act of 1992 forced all ESAs to corporatize their operations, moving to a commercial company structure with shareholders and profit objectives. In the case of municipally owned networks this was a straightforward process, since they had well-defined owners and already operated on a commercial footing. In the case of the rural Electric Power Boards (EPBs), however, no defined owners existed. The Boards had been set up from 1918 on as “creatures of statute” which installed and managed their network assets on behalf of the consumers who elected the boards. Under corporatization, EPBs were deemed to be owned by all consumers served at the moment of the changeover. A variety of creative schemes for issuing shares were implemented in the early 1990s. Some Boards, transformed into joint-stock companies, issued shares to newly created elected trusts which held the shares on behalf of consumers in the same way as the EPBs had previous held their real assets. In other cases shares were gifted to individual consumers, many of whom took the opportunity to cash in by selling shares to private-sector interests, which quickly aggregated them into sizeable voting blocs. A period of consolidation by mergers and takeovers followed, as the more entrepreneurial of the new companies bought-up shares where possible, or took over control of trust-owned companies by direct acquisition where trust boards were willing. By early 2003, the four largest companies had captured 60%

Restructuring of the New Zealand Electricity sector 1984–2005

217

Table 7.4. Consolidation of market shares in distribution networks: GWh carried.

Power New Zealand/United Networks Vector Ltd Powerco Orion Ltd Total, big four Other companies Total GWh Share of big four (%)

1995

1998

2001

2004

2569 4053 347 2416 9385 13,700 23,085 40.7

3384 4432 1019 2582 11,418 14,422 25,840 44.2

7120 4990 2083 2822 17,015 10,711 27,726 61.4

** 10,257 4074 3080 17,412 12,488* 29,900* 58.2

*Estimate. **Taken over 2003 by Vector and Powerco, who divided up the network assets. Source: MED disclosure statistics, at http://www.med.govt.nz/ers/inf_disc/disclosure-statistics/, plus company disclosures for 2004 financial year.

of the distribution lines business, up from 40% 10 years earlier. In 2003 a further merger reduced the number of industry leaders to three (see Table 7.4).25 The Electricity Industry Reform Act of 1998 forced ownership separation of electricity retailing from the operation of distribution networks. Most of the existing distributors opted to retain their natural-monopoly lines businesses and divest their retail arms. The retail businesses, with their customer bases, were quickly snapped-up in 1999– 2000 by the five main generators, which thereby achieved vertical integration of their generation plants with retail outlets.26 The supply of wholesale power to these retail affiliates then became an intra-firm transfer, largely removing any need for the large generators to enter into openmarket long-term contracts or sell more than a marginal part of their generation through the spot market. In the very light-handed New Zealand regulatory environment of the 1990s, vertically integrated generator retailers had a strong competitive advantage over stand-alone retail

25

It appeared to some observers in the 1990s that the new corporate culture of the major network companies, with its focus on mergers and acquisitions, might shift management priorities from ensuring reliability of supply to financial issues such as the market valuation of the enterprises. Claims of this sort were heard especially in relation to the failure of all the high-tension cables supplying the downtown Auckland area in 1998, due to a combination of improper installation and poor maintenance practice. An inquiry into the failure concluded that “Mercury (the relevant network company, since renamed Vector Ltd) does not have an adequate maintenance policy for 110 kV gas and oil filled cables. It did not comply with manufacturers’ recommendations in regard to the routine testing of gas pressure and oil pressure alarms and accuracy of their initiating devices, and electrical checking of the integrity of the outer coverings of the cables.” See Integral Energy Australia, Inquiry into the Auckland Power Supply Failure http://www.med.govt.nz/inquiry/publicsummary.html#P117_7323 conclusion xvii.) These failings, however, predated the corporatization process and at most it would seem that the new culture failed to remedy them. 26 Non-major retailers survived only in a few isolated rural areas such as the King Country in the central North Island. (King Country Energy’s independent-retailer status is buttressed by ownership of (and vertical integration with) local small hydro amounting to 50% of its retail load. It also has a 50% share in the large Mangahao hydro station in the Manawatu. See http://www.kcenergy.co.nz/ generation.html

218

Electricity Market Reform

businesses because of their ability to hold physical hedges27 within each company, whereas independent retailers had either to secure hedge contracts from generators on an extremely thin market, or face exposure to the spot price. Even faced with a dry-year crisis in 2001, the New Zealand Government took no steps to compel generators to offer hedge contracts on the open market. With no regulatory or statutory protection against the exercise of market power by the vertically integrated generator-retailers, almost all independent retailers were deprived of either profitable arbitrage opportunities or access to profitable long-term contracts, and quickly exited the market. Only a single large independent retailer remained by the end of 2000. In 1996 the Canadian company TransAlta had acquired a substantial share of the distribution and retail market, but in 1999 the company was unable to acquire a large enough generation portfolio to match its retail sales volume.28 Faced with large upstream exposure to the hedge and spot markets, TransAlta quickly sold its New Zealand business for $830 million29 to New Zealand’s dominant natural-gas pipeline and retail company, NGC. Possessing only 399 MW of generation capacity, and having failed to secure forward hedge contracts to cover the winter of 2001, NGC’s retail affiliate On Energy found itself in June 2001 in a critically dry winter with almost full exposure to the spot market for its supply of electricity.30 The company could not raise its retail price to cover the high wholesale prices, because its vertically integrated competitors kept their retail prices unchanged throughout the crisis. As NGC’s subsequent annual report ruefully noted, recording losses of $304 million from this classic cost–price squeeze:31 “Wholesale prices increased to up to four times their normal levels, placing a pronounced strain on NGC’s cash flows, profitability and financing arrangements, and raising serious questions about the operation of the market itself. NGC decided to withdraw from electricity retailing and completed its exit on August 1, 2001 following the sale of its retail electricity customers to two Government-owned energy companies. NGC’s withdrawal from that business closed off future retail exposure to the volatile wholesale electricity market and crystallized the resulting losses.” Of the retail customer base of 405,000 which NGC had acquired from TransAlta NZ Ltd the previous year, representing 23% of all electricity consumers, 115,000 were sold to Meridian Energy and 290,000 to Genesis Power Ltd. Since then the vertically integrated five generator oligopoly of retailing has been unchallenged. The elimination of non-generator parties from the retail market spelt a halt to the process of competition for retail customers, which had briefly flourished in the 2 years following the 1998 separation of lines and energy retail activities. Figure 7.5 shows that the new-entrant

27

Retailers can hedge their costs of future wholesale supply either by long-term contracts with generators, or by directly owning physical generating plant. The practice of physical hedging in New Zealand has foreclosed the emergence of a liquid hedge market; this in turn has constituted a major barrier to new entry by independent retailers. 28 TransAlta in 2000 held more than 20% of New Zealand’s electricity consumers but less then 5% of generating capacity. 29 NGC Becomes Majority Owner of TransAlta, media release dated 31 March 2000, http://www.ngc.co.nz/ article/articleprint/166/-1/21/. The price represented a $300 million tax-free capital gain for TransAlta. 30 The wholesale spot-price spike of June–August 2001 is dramatically apparent in Figure 7.3 above. 31 Natural Gas Corporation, Annual Report 2001, p. 5.

Restructuring of the New Zealand Electricity sector 1984–2005

219

40 35 30 (%)

25 20 15 10 5 April 1994 October 1994 April 1995 October 1995 April 1996 October 1996 April 1997 October 1997 April 1998 October 1998 April 1999 October 1999 April 2000 October 2000 April 2001 October 2001 April 2002 October 2002 April 2003 October 2003 April 2004 October 2004

0

Fig. 7.5. Share of non-incumbent retailers in former franchise territories. Source: Stratagen.

share of retail sales in former franchise territories, following removal of franchises in 1993–1994, remained very low until the Electricity Industry Reform Act 1998 separated retail from distribution. Retail competition took off in 1999–2000, but froze again at around 30% as soon as On Energy had been driven out in mid-2001. Three years later, Murray and Stevenson (2004, p. 18) reported to the Electricity Commission that “customer switching figures seem to have declined and stabilized over a period when prices have been rising” and that “price trends suggest electricity prices are probably higher on average than they would be in a workably competitive market”. The 1989 Task Force vision of competitive retail markets served by a liquid market for forward hedge contracts, thus, ran aground on the reality of generators’ market power. The anti-competitive effect of vertical integration of generation with retail had not been foreseen at the time of the 1998 separation of retail from distribution networks. Consequently no consideration was given to requiring generators to transact with their retail affiliates via an arms-length contestable market for hedge contracts, and although proposals for such compulsory hedging were discussed during the 2001 crisis, Government took no steps to remedy the extreme thinness of the forward contracts market.32

7.4. Pricing, Profitability, and “Light-Handed Regulation” 7.4.1. “Efficient” pricing A dilemma over the meaning of “efficient prices” dogged the electricity reform process from the outset, and remains an unresolved issue two decades later. One interpretation in the mid-late 1980s was that since the electricity system was breaking even in cash terms at its existing prices, efficiency-enhancing reforms ought to bring down the prices paid by final consumers, and certainly ought not to lead to rising prices.

32

The issue now rests with the recently established regulator, the Electricity Commission.

220

Electricity Market Reform

A May 1990 press statement by the then Minister of Energy reassured consumers:33 “Lower real electricity prices resulting from the corporatization of the electricity distribution industry is the motivation for the latest Government decisions on electricity which were announced today … Savings in electricity bills represent an immediate improvement in living standards and help toward the restoration of full employment ….” An opposing view from the outset was that economic efficiency required prices to increase, to raise the industry’s return on capital to a commercial level (Ministry of Energy Financial Objectives and Pricing Review Team 1984). In addition, an across-the-board price increase was allegedly needed to signal to electricity users the marginal cost of new supply (Electricity Corporation Establishment Board, 1987; New Zealand Treasury, 1987). There was general agreement that low-cost generation options had been fully exploited by the 1980s, and that new generation and transmission capacity would be costly to install. Faced with an upward-sloping LRMC curve, the choice between average- and marginalcost pricing presented a political dilemma. If the restructured electricity industry were to be allowed to price at LRMC, the inevitable result would be higher prices to consumers and very large operating surpluses on the existing hydro generation plant, far in excess of the surpluses required to yield a competitive return on, and of, the book value of alreadyexisting capital (Bertram, 1988). If a lower average price were set to recover the full cost of supply, including a commercial rate of return on the book value of existing assets, then the resulting price signal would render new investments unattractive while encouraging excessive growth of demand. Two solutions to this dilemma were on offer. The consumer-oriented position was either to stick with an average-cost price and accept any consequent inefficiencies;34 or to adopt a non-linear tariff structure to achieve the same outcome of restricting existing generators’ total revenue, while providing efficient price signals at the margin. The latter solution was supported by a parliamentary select committee (New Zealand House of Representatives, 1992) and in a report commissioned by a group of major users (Terry et al., 1992).35 The other approach to wholesale pricing, championed by the Treasury and ECNZ, was to charge consumers the full LRMC price, and to legitimize the resulting cash surpluses that would accrue to generators, the grid operator, and the distribution networks, by revaluing their existing assets up to a level at which the rate of return on capital would appear to be no more than “normal”. In 1987 Treasury had estimated that the BST should be raised from less than 6 c/kWh to somewhere in the range 8–11 c/kWh (New Zealand Treasury, 1987, p. 4). The Labor Government, which initiated the reforms, was replaced at the 1990 general election by a National Party regime in which the Treasury view prevailed. In terms of

33

Hon David Butcher, press statement dated May 25, 1990. Advocates of this approach in the mid-1980s included the New Zealand Business Round Table (1985), Ernst and Whinney (1985), Frater et al. (1985) Jarden and Company (1985), McDonald (1985), Scott and Co (1985), and University of Waikato Interfirm Comparison Unit (1985). 35 Another pricing arrangement with the same basic thrust would have been to rebate to consumers any excess profits resulting from application of a uniform LRMC price, possibly by means of a lump-sum reduction in fixed lines charges funded from generation surpluses, along the lines later adopted in the UK by Scottish Hydro. 34

Restructuring of the New Zealand Electricity sector 1984–2005

221

electricity-sector reform, this meant support for “full-cost uniform pricing” of electricity, which translated in practice into higher overall prices for consumers, with any efficiency gains that might result from restructuring being captured as additional profit. Treasury argued that electricity prices needed to rise rather than fall, to signal LRMC; that no regulatory barrier should be placed in the way of electricity suppliers pushing their prices up to the “limit prices” at which, in theory, the threat of entry by new competitors would cap prices; and that gains from increased prices and/or reduced costs, provided they fell below the contestability threshold, could legitimately be taken as profits and built into the asset valuations shown in the companies’ regulatory accounts. This tolerance for wealth transfers from consumers to suppliers36 meant that New Zealand’s regime of so-called “light-handed regulation” lacked any bright-line test for abuse of market power until all assets had been revalued up to the replacement-cost ceiling, and companies had adjusted their margins to match the higher ratebase. It also reveals the extent to which New Zealand policy-makers adopted without qualification some recent developments in economic and accounting theory, which other OECD governments have treated with more circumspection. 7.4.2. Economic and accounting theory and the New Zealand reforms Economic policy-making in New Zealand in the late 1980s and early 1990s was heavily influenced by three overseas developments in the economics and accountancy literature. These were: ●

●

●

36

The proposition, familiar from early UK debates over electricity restructuring, that electricity generation and retailing were potentially competitive activities and that in relation to those two levels of the electricity market, therefore, policy intervention could be limited to promoting competitive conditions, not to controlling prices. The theory of contestable markets set out in Baumol et al. (1982). Contestability theory was interpreted to mean that in a process of “competition for the market”, a natural monopolist would be unable to price above the limit at which a new entrant would be attracted. This, New Zealand officials reasoned, meant that if an incumbent monopolist’s assets were revalued up to replacement cost, no more than a competitive rate of return on that valuation would be achievable unless management could cut costs by improving efficiency. Hence, although electricity lines networks were natural monopolies, officials decided no regulatory restraint on price would be necessary, as market disciplines would do the job unaided; all that would be required would be transparent information disclosure. The newly fashionable method of accrual (current-cost) accounting, which prescribed that fixed assets should be continually revalued to market value, and that profit and loss statements ought to reflect changes in shareholder wealth accruing as a result of each year’s trading activity. In the hands of the New Zealand accounting profession, this methodology was incorporated into “generally accepted accounting practice” (GAAP) in a partial manner that opened the way to manipulation of asset valuations. To summarize a complex story, New Zealand’s Accounting Standard SSAP28 (later FRS3) prescribed that natural-monopoly entities whose assets do not (by definition) have a competitive

The two Government departments most closely associated with electricity regulation during the 1990s, Treasury and the Ministry of Commerce, adopted and promoted the so-called “total surplus standard” for regulation. This standard treats all pure transfers as welfare-neutral and hence of no concern to the regulatory authorities. See Bertram (2004).

222

Electricity Market Reform

arms-length market value, should value their fixed assets at optimized depreciated replacement cost (ODRC), which was claimed to approximate the capital cost of setting up from scratch a new supplier providing the same service, to the same standard, as the incumbent (Cooper, 1995). This valuation could then be used as the ratebase for setting and justifying prices. In the view of the officials overseeing the light-handed regulatory regime during the 1990s, no concern over excess profits could arise so long as no more than a competitive rate of return on the ODRC-valued assets was revealed in the regulatory accounts prepared for disclosure purposes by all transmission and distribution network owners. If assets were to be continually revalued to the hypothetical contestability limit, consistency required that the profit-and-loss account should record as income all wealth changes accruing to the shareholders, whether by virtue of current cash flows or of asset revaluations. New Zealand’s GAAP, however, did not (and still does not) require this to be done for upward revaluations. Gains and losses on the actual sale of particular assets are recorded in the profit-and-loss account, as are all negative revaluations (asset write-downs). The crucial omission is the treatment of upward asset revaluations (effectively, negative depreciation). Rather than being recorded as revenue in the profit-and-loss account, these are recorded separately in a “revaluation reserve”, usually hidden deep in the notes to the financial statements. Under this procedure, the accrual to a company’s books of hundreds of millions of dollars of revaluations of fixed assets need never be recognized as income, and so can be excluded from recorded profits for both taxation and regulatory purposes,37 while the revalued assets can be used as the ratebase for price setting and justification. 7.4.3. Generation and the wholesale spot price Figure 7.6 shows the generation supply curve for May 2004, constructed by stacking the various generation plants in merit order of variable operating cost. The large hydro plants, 11 9

c/kwh

7 5

Average monthly price ⫽ marginal cost of last plant in the stack

3

2004 supply curve May demand

1 -1 380

880

1380

1880 2380 GWh per month

2880

3380

Fig. 7.6. Generation supply curve 2004.

37

This practice is acceptable to the tax authorities because New Zealand does not have a capital gains tax.

Restructuring of the New Zealand Electricity sector 1984–2005

223

whose operating cost is close to zero, crowd the higher-operating-cost thermal and geothermal units out to the marginal one-third of the market. The upward-sloping curve at the right of the diagram shows these various non-renewable units stacked in merit order. The market-clearing spot price, which provides, over the long run, the anchor for long-term wholesale supply contracts, is found at the point on the supply curve at which aggregate demand intersects the supply curve. The May 2004 demand, it can be seen, lay only about 400 GWh (14% of monthly demand) inside the point at which the supply curve turns sharply upwards. In this situation, radical price spikes can be anticipated if either demand rises, or supply falls, by this amount. The months following May are winter in New Zealand, when demand is higher and the system’s ability to meet demand without price shocks rests heavily on the volume of very low-operating-cost hydro generation made available by the owners of hydro plant. By withholding even a small part of the available water from use for generation at times of strong demand, the owners of large hydro plants can potentially pull the bidstack to the left, thereby (deliberately or inadvertently) driving up the spot price and raising their operating surplus – an opportunity for the exercise of market power mitigated in the New Zealand case only by the existence of a duopoly, rather than a monopoly, of major hydro generators with the necessary market leverage. The very steep profile of the supply curve beyond about 3000 GWh per month confers substantial market power on any hydro generator (or cartel of generators) with the ability to withhold capacity and thereby shift the bidstack to the left. To achieve such withholding, a hydro generator must either have unutilized water storage capacity which can be allowed to fill while generation is curtailed; or else must be able to dispose of unwanted water by hydro spill. New Zealand policy-makers became aware only in 2001 (5 years after the break-up of the ECNZ generation portfolio) of the possibility that hydro generators might game the spot price by spilling water to waste. The Government’s ex-post review of the 2001 dry-winter supply crisis brought to light the fact that in the summer of that year Meridian Energy had been spilling water from Lake Tekapo. Whether this was strategic behavior to drive up price (as one distributor alleged), or responsible management to avoid flood risk (as Meridian claimed),38 the issue was placed on the agenda for regulation, and new rules subsequently came into force requiring generators to report each month on the details of any spill.39 Since 2001 there has been very little hydro spill recorded. Use of empty storage capacity to withhold water, however, is not so subject to Government control. An example of the strategic importance of commercial generators’ restriction of hydro generation in order to build up (or protect) the level of storage lakes was the price spike of April 2003, visible in Figure 7.3. Rainfall in the early months of 2003 was below normal, and lake storage fell below the levels required to ensure ability to meet the forthcoming winter demand. The two large hydro generators in the South Island (Meridian and Contact) both cut back water use, citing the need to conserve water and maintain storage levels ahead of the coming winter. At the same time, Contact took its 357 MW gas-fired Stratford station offline in mid-April for

38 See Electricity Post-Winter Review, 2001, Section 2.2, http://www.winterreview.govt.nz/submissions/ summary/summary-03.html#P186_28826 39 Hydro spill reporting is now to the recently established Electricity Commission; see http://www. electricitycommission.govt.nz/opdev/secsupply/sos/overview/hydrospill1/

224

Electricity Market Reform

maintenance.40 Both actions shifted the bidstack significantly to the left. As lake storage levels dropped to 60% of normal for the time of year, the spot price was driven up sharply to an average for the month of 20 c/kWh, and Government solicited voluntary demand restraint by electricity users in order to avoid blackouts. Rainfall subsequently increased during May and June and the supply situation eased, bringing the spot price back down to 6 c/kWh by June. The extreme volatility of the spot market in the April 2003 event was attributable not only to the high shadow price implicitly assigned to water by Contact and Meridian Energy. It was worsened significantly by the fact that the New Zealand bidstack in 2003 had far less reserve thermal plant, and hence a much steeper right-hand end, than had been the case at the beginning of the reforms. In the dry year 1992 the ECNZ portfolio had included four high-operating-cost thermal plants, which were brought online to compensate for the water shortage. Because these plants’ capital costs were sunk, the only economic cost of bringing them online was the operating cost, primarily fuel. By their mere existence, these plants exercised a moderating influence over the spot market by placing a ceiling on the spot price over a range of several hundred MW of supply capacity. Table 7.5 shows the thermal high-cost reserve capacity that had been available in 1992 (when operation of the Marsden A oil-fired station during the dry-winter crisis reduced the scale of blackouts in the Auckland region), and compares this with the corresponding reserve capacity available in early 2003 before the April price spike. The difference is striking. Under commercial incentives and supposedly competitive conditions, the former owners of thermal reserve plant had decommissioned and/or demolished a total of 620 MW of reserve capacity.41 Over the same period, roughly 1000 MW of new thermal plant was commissioned, but none of this qualified as reserve capacity to cover dry years; the Southdown and Otahuhu B stations simply helped supply to keep up with growing demand, while the cogeneration stations perform no role in relation to dry-year firming, since their operation is tied to the steam requirements of the host facilities. Having failed to persuade any of the commercial generators to invest in new reserve plant, the Government opted in 2004 to spend $160 million on construction of a new 155 MW diesel-fired thermal station at Whirinaki, where Contact Energy had demolished an almost identical plant a couple of years previously. The station, although owned by the Crown, is maintained and operated by Contact Energy under contract, and is not to be dispatched at a price of less than 20 c/kWh42 (roughly the monthly average price during the April 2003 price spike, see Fig. 7.3). 7.4.4. Grid pricing The high-voltage transmission grid was transferred in 1994 to a new State-Owned Enterprise, Transpower Ltd, following several years of debate over asset valuation and pricing. 40

NZEM, Declining Storage Levels Fuel Rising Electricity Prices 7 May 2003, http://www.nzelectricity.co.nz/ C2dPricesMonth/030508.htm 41 Prior to 1992, the 133 MW Meremere coal-fired station in the Waikato had already been decommissioned by ECNZ in 1990. Marsden A (114 MW) was closed in mid-1992 and demolished in 1997. Stratford (200 MW) closed in late 1999. Otahuhu A (90 MW) and Whirinaki (216 MW) were decommissioned in 2002. 42 Electricity Commission, Explanatory Paper to the Initial Security of Supply Policy, June 2005, http:// www.electricitycommission.govt.nz/pdfs/opdev/secsupply/policy/Initial-SOS-Policy-ExplanPaper.pdf, Part VII p. 21.

225

Restructuring of the New Zealand Electricity sector 1984–2005 Table 7.5. Thermal generating capacity, March 1992 and March 2003 compared. Station New Plymouth Huntly Stratford TCC Otahuhu B Southdown Big thermal total Stratford Otahuhu A Marsden A Whirinaki Total high-cost dry-year reserve thermal

1992 capacity (MW)

Operating cost, c/kWh, 1991

2003 capacity (MW)

580 1000 198 0 0 1778

3.13 2.92 3.97 n.a. n.a.

400 1000 355 380 118 2253

200 90 114 216 620

3.97 6.27 7.43 18.5

0 0 0 0 0

0 0 0 0 0 0 0

n.a. n.a. n.a. n.a. n.a. n.a.

52 40 44 25 355 65 581

Te Awamutu cogen Kinleith Te Rapa Edgecumbe Kapuni Whareroa Cogen total Total thermal

2398

2834

Sources: 1992 capacity data from Annual Statistics in Relation to Electric Power Operation in New Zealand for the Year Ended March 31, 1992, pp. 57–59. 2003 capacities from Energy Data File July 2003, pp. 108–109. Operating-cost estimates from Terry et al. (1992), p. 128.

Following the 1987 transfer of the NZED generation and grid assets to ECNZ at a negotiated vesting value of $6.3 billion, ECNZ undertook the task of allocating this lump-sum valuation across its generation and grid assets. The transmission system was assigned a value of $2.1 billion, and generation and other fixed assets $4.2 billion.43 In July 1990 the Transpower Establishment Board was set up to oversee the separation of the grid from ECNZ. A central issue confronted by the Board was the valuation that should be assigned to the grid assets when they were fully vested in a new independent company. ECNZ management and Treasury were focused on achieving privatization of the generation assets at a high price, and this could best be achieved by off-loading as much as possible of the Corporation’s debt into the books of its grid subsidiary, allowing the generation assets to be sold relatively unencumbered by debt. In addition, a range of operating expenses formerly attributed to generation were transferred to Transpower prior to separation (Terry et al., 1992, p. 87), raising the reported profitability of ECNZ’s generation business in readiness for sale. A higher valuation of the grid assets was then required to bring Transpower’s debt– equity ratio down to a commercially sustainable level. The TPEB achieved this objective by having the grid assets revalued to “optimized deprival value” (ODV), a variant of depreciated replacement cost. This resulted in a valuation of $2.55 billion (Ernst et al., 1991). The higher asset value and increased operating costs were used to justify a real increase of 21% between 1989 and 1991 in the grid transmission charge per kWh conveyed. 43

ECNZ Annual Report 1989, p. 47.

226

Electricity Market Reform

Over the decade following its establishment as an state-owned enterprise, Transpower paid down its debt and wrote-down its ODV asset valuation in recognition that the longrun sustainability of the grid itself depended upon transmission prices low enough to compete with distributed generation connected directly to distribution networks, for which transmission service would not be required. To protect the grid’s pre-eminent position in the short term, Transpower used its market power to impose contract conditions on distributors which obliged them to collect transmission charges on all power delivered, whether it was taken from the grid or generated locally by suppliers connected only to the local network. These contract conditions, by imposing high fixed connection charges regardless of load changes, also eliminated the prospect that retailers might be able to profit from demand-side conservation initiatives. The resulting barrier against entry by small-scale distributed local generation, and the equally suffocating effect on local demand-side conservation initiatives, effectively foreclosed development of both for a decade. 7.4.5. Distribution networks Legislation to force through the corporatization of ESAs was passed in 1992, and the process was largely completed by April 1994. As the new companies were set up, the issues of asset valuation and price setting had again to be addressed. Following the Transpower precedent, the Minister of Energy and the Treasury planned to revalue all assets up to ODV prior to vesting, enabling the new distribution companies to start off with a new, higher ratebase against which their profitability could be monitored under a light-handed regulatory regime of information disclosure. It was obvious to all industry participants, including major users, that the historic-cost asset valuations in the books of the pre-corporatization ESAs were far below depreciated replacement cost. Roughly speaking, at 1994 the network assets of all networks combined had a book value of $2 billion, but a replacement-cost valuation would come to double that amount.44 If the new companies were gifted a $2 billion asset revaluation at the time the assets were vested, two politically significant groups stood to lose. One group was electricity users, who effectively would have to pay for the increased profits required if the distribution companies were to meet commercial rate-of-return targets on their revalued ratebases. The other group were private investors eager to make capital gains by acquiring distribution assets cheaply and then undertaking the revaluations themselves. Early in the restructuring process it became apparent that switching to a replacement-cost ratebase for pricing supply to customers in low-density areas would sharply increase electricity prices in low-density rural areas with a high ratio of line length per customer. A confidential survey undertaken by officials in 1989–1990 found that “full-cost pricing” would require price increases of up to 300% for rural electricity users.45 Faced with the prospect that the political fallout would halt the reform process at the outset, Treasury fell back on a modified form of replacement-cost valuation called ODV, which included the proviso that whenever the economic value of an asset (the discounted present value of expected revenues46) was below full ODRC, the asset would be written down and the users of the asset 44

Cabinet documents recently released under the Official Information Act reveal that these orders of magnitude were known to ministers and officials in 1991, 3 years before vesting took place. 45 Cabinet committee document SAS (90) 31, March 13, 1990, p. 10. 46 The circularity between asset values and revenues was well understood. The ODV technique enabled the revenues extracted from specific groups of consumers to be selectively capped, with the ratebase valuation of the assets serving that group written down accordingly, leaving an ostensibly competitive market return on the assets for disclosure purposes.

New Zealand $ billions

Restructuring of the New Zealand Electricity sector 1984–2005 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

227

Revaluations Depreciatedhistoric-cost estimate 1992 1994 1996 1998 2000 2002

Fig. 7.7. Asset book values of electricity distribution networks, 1992–2002. Source: Bertram and Terry (2000, p. 7).

thereby protected from rate shock. This ingenious solution became embedded thereafter in the valuation procedures for both grid and distribution networks. In the event the new National Government elected in late opted pragmatically to continue the time-honored practice of using revenues from densely populated parts of each ESA’s territory to cross-subsidize the prices charged in low-density areas – a procedure which had been generally accepted by consumers since the 1920s.47 Neither the ODV concept nor the decision to retain urban–rural cross-subsidies removed the looming prospect of a general price shock if network asset valuations were doubled across the board. In October 1991, Ministry of Commerce officials estimated that the ODV valuations would be 2.5 times the existing book values.48 Modeling carried out for the Government in April 1992 by a local accountancy practice suggested that a rate shock of 25% would be required to meet the required return on a revalued ratebase.49 Treasury at this stage proposed that the assets should be vested at book value but that the new companies be allowed to revalue to ODV without facing any regulatory restraint. It would then be the responsibility of the new corporate boards to decide whether to squeeze their customers or accept below commercial rates of return.50 Cabinet agreed,51 and the Establishment Boards of the new companies were instructed to adopt existing book values for their opening balance sheets. Figure 7.7 shows the subsequent process of increasing the regulatory ratebase by writing-up asset values to replacement cost. Figure 7.8 shows the evolution of prices and average costs of lines networks over that period. Free from regulatory restraint, the sector raised its aggregate Lerner Index from 0.36 at vesting to 0.68 by 2001. The loophole in the regulatory system was well known to, and understood by, industry insiders. It was equally obvious to analysts familiar with current-cost accounting theory. The procedure of vesting the assets at historic cost, while signaling to the new owners that ODV valuation would be the regulatory benchmark, transferred responsibility for 47

Corporatized ESAs are compelled, under the reform legislation, to maintain supply to all rural customers until 2013. Thereafter they will be allowed to disconnect unprofitable customers. 48 Ernst and Young, letter to Michael Lear, Ministry of Commerce, May 14, 1992, p. 1. 49 Ibid., p. 3 of appendix. Ernst and Young pointed out in this letter that recognizing asset revaluations as income would reduce the required rate shock to between 5% and 9%, but the point was not taken by officials. 50 Officials’ briefing document for Minister of Energy, May 8, 1992. 51 Cabinet State Sector Committee document STA (92) 96.

228

Electricity Market Reform 5.00 4.50 c/kWh conveyed

4.00 3.50 3.00 2.50 2.00

Average revenue

1.50

Average operating costs

1.00 0.50 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.00

Years to March Fig. 7.8. Price–cost margin of electricity distribution networks, 1991–2002. Source: Bertram and Twaddle (2005, p. 295), Figure 1(f).

subsequent increases in margins and prices from the Government to the distributors but left consumers unprotected. In a current-cost accounting framework, profitability should be measured with revaluations (changes in shareholder wealth) recorded in the profit-and-loss accounts. New Zealand’s GAAP did not require this to be done.52 Lines companies were therefore able to inflate the denominator and reduce the numerator in their profit calculations, as justification for an annual wealth transfer from consumers to distribution network owners of $200 million annually (0.2% of GDP) from the late 1990s on (Bertram and Twaddle, 2005). Perhaps ironically, the information disclosure regulations for electricity lines networks, promulgated in 1994, included a requirement for companies to disclose an “accounting rate of profit” which included the wealth effects of ratebase revaulations,53 and this requirement was complied with, resulting in the disclosure of profit rates often of 30–40%, and in one case as high as 90%,54 with no reaction from Government.55 52

This issue had been thoroughly discussed prior to the UK privatizations, and the regulatory accounting implications worked out, in the “Byatt Report”, Accounting for Economic Costs and Changing Prices: A Report to HM Treasury by an Advisory Group, London: HMSO, 1986, Volume 1. 53 Ernst and Young, as advisers to the Ministry of Commerce, set out the correct accounting procedures in a letter of May 14, 1992, and explained the correct interpretation of the Accounting Rate of Profit (later renamed the Return on Investment) in Ernst and Young (1994). 54 Far from recognizing the implications of these numbers, a 2000 Ministerial Inquiry rejected the calculation methodology itself and found no grounds for regulatory concern (Caygill et al., 2000, Table 7.3, p. 14, and p. 15 paragraph 75). 55 The largest lines company, United Networks, disclosed a return on equity of 235% for 2000, 347% for 2001 and 125% for 2002, without attracting attention from Parliament, media, or officials. See New Zealand Gazette 2000, No. 111 p. 2807 (http://www.dia.govt.nz/Pubforms.NSF/URL/UnitedNetwork 111Aug00.pdf/$file/UnitedNetwork111Aug00.pdf ); 2001, No. 104 p. 2665. (http://www.dia.govt.nz/ Pubforms.nsf/URL/Unitednetworks104Aug01.pdf/$file/Unitednetworks104Aug01.pdf); and 2002, No. 122 p. 3272. (http://www.dia.govt.nz/Pubforms.nsf/URL/UnitedNetwork122.pdf/$file/United Network122.pdf ). In fairness it should be noted that the taking of monopoly profits is not illegal under New Zealand competition law. Consumers have no legal redress against high prices, and the Electricity Complaints Commission set up in 2001 was barred from hearing complaints about pricing. See http:// www.electricitycomplaints.co.nz/faqs.htm

Restructuring of the New Zealand Electricity sector 1984–2005

Contestability limit

Market value

229

Say 2.5 ⫻ ODV ⫽ $8.4 billion

$ billions

Full replacement cost

Depreciated replacement cost

Costless-entry limit

Optimised DRC/ODV

4.2 billion

Historic cost

$2 billion

Net realisable value 0 Fig. 7.9. Range of possible ratebase valuations for the distribution networks.

Figure 7.9 shows schematically the range of feasible asset valuations, any of which could have been arbitrarily chosen for ratebase purposes. The theoretical limit valuation under conditions of perfect contestability (zero costs of entry and exit) is represented by the ODV of $4.2 billion – more than double the pre-corporatization historic cost. Adding in the observed effects of barriers to entry (in particular, very high fixed costs of entry and exit) raises this further by a factor of 2.5 (based on the actual purchase price of distribution networks taken over as going concerns). In short, the New Zealand regulatory regime for lines businesses prior to 2003 encouraged ratebase revaluation up to ODV, which was achieved by the network businesses over the first 6 years of reform. Thereafter, as market expectations factored in the lack of credibility of the light-handed regime, network assets changed hands at “fair value” levels, which included the discounted value of expected future regulatory tolerance. The market judgment in these transactions suggested that an actual contestability limit valuation would be of the order of $8.4 billion for all networks aggregated. The essential issue raised by asset revaluations throughout the electricity sector was not the theoretical choice of valuation methodology per se; there are ample precedents around the world for both the historic-cost and the replacement-cost approach, with matching implications for the setting of the warranted rate of return on the resulting ratebase. The central issue was the New Zealand Government’s decision to radically change the ratebase valuation methodology in asset mid-life, causing a dramatic levy (several billions of dollars) on the aggregate wealth of consumers, for the benefit of electricity suppliers. No protection was provided for consumers against this wealth expropriation. In particular, no regulatory provision required suppliers to compensate consumers for the wealth transfer, whether by means of rebates or through allocation of shares in the newly created equity value of suppliers. In August 2001 Parliament passed a set of amendments to the Commerce Act 1986, giving the New Zealand Commerce Commission the task of regulating transmission and distribution lines networks. The Commission conducted lengthy hearings on the pricing practices of the electricity networks sector, and eventually decided to use the status quo of mid-2002 as its ratebase for future profit-cap regulation. The revaluations and widening margins of

230

Electricity Market Reform

the 1990s were thereby retrospectively legitimized.56 This regulatory outcome was inherited in 2003 by the new Electricity Commission.

7.5. The Electricity Commission: Back to An Industry Regulator After more than a decade of experimentation with light-handed regulation, the New Zealand Government finally in 2003 followed the example of most other OECD countries by setting up a specialized industry regulator to oversee the electricity industry. The Electricity Commission is charged with a wide array of tasks: managing the ongoing information disclosure regime, setting price and revenue caps, coordinating the investment plans of various industry players, maintaining reserve generation capacity, overseeing industry governance arrangements, and guiding new investment by the issuing of “statements of opportunity”, to name a few. Further privatization is off the policy agenda. Several regulatory issues, however, remain unresolved: ●

●

●

There is little prospect that the incumbent generators will be forced to divest their retail affiliates; yet without such divestment, new competitive retail entry remains foreclosed. Similarly, although Government has declared itself in favor of the rapid development of distributed generation, Transpower’s grid pricing practices, which foreclose most opportunities for such projects, remain in place. Since 2002 a rush by large incumbent generators to build wind farms is raising a raft of difficult coordination problems, since the location of favorable sites for wind farms, and of the hydro generation assets that can be used to back-up wind generators, does not always coincide with the existing grid infrastructure, presenting the grid’s operator, and the new regulator, with investment and coordination requirements not foreseen even a few years ago.

7.6. Conclusion: The State of Play at 2005 The structure of the industry in 2005 is shown in Figure 7.10. Of the 1989 Task Force recommendations, some have been implemented while others have been abandoned along the way. ECNZ has been broken into five separate generators (the Task Force had recommended against breakup). Only two of these generation companies are in private hands, while the Government continues to own 60% of generating capacity. The Task Force’s fear that generation breakup without an industry regulator might result in losses of efficiency in the coordination of scheduling and investment seemed to have been borne out by 2002, and partly in response to this a new industry regulator was introduced in 2003. Generation and transmission were separated early in the reform process, but the Task Force’s proposal for club ownership of Transpower was rejected early on by industry participants, leaving the grid in state ownership.57

56

The Commission’s deliberations are fully recorded at http://www.comcom.govt.nz/Industry Regulation/Electricity/ElectricityLinesBusinesses/Overview.aspx 57 The Government attempted to implement the club proposal in 1992, but distributors refused to take part in the formation of a club in which their interests would have been diametrically opposed to those of generators, but in which they would not have had sufficient voting power to form a blocking coalition. The Task Force had failed to appreciate the likely extent of these conflicts of interest.

231

Restructuring of the New Zealand Electricity sector 1984–2005 Share of generation Generation

Electricity commission

26.9%

31.2%

16.6%

12.6%

7.9%

4.8%

Contact Energy

Meridian Energy

Genesis Power

Mighty river Power

Other IPPs

Cogen

Electricity Governance Regulations and Rules 2003

Transpower

Transmission

Local network companies

Distribution

Electricity retailers

Retailing

Consumers

Large directsupply industrial consumers

Residential, commercial, agricultural, small and medium industrial consumers

Fig. 7.10. Electricity industry structure 2004.

Corporatization of ESAs has been carried through, but less than half of distribution network assets have been fully privatized, and the interim trust-ownership arrangement has become entrenched in many rural and small-town systems. Retail franchises have been abolished and retail operators separated from lines networks, but competition at retail level quickly stalled once retailers and generators became vertically integrated. No liquid market for hedge contracts has yet emerged – a defect still to be addressed by the new industry regulator. The main buyers in the wholesale market are the retail affiliates of generating companies, plus major manufacturers taking supply directly from the grid. Direct consumer exposure to spot market prices was estimated in 2003 to be no more than 10–15%,58 which is not surprising given that the great bulk of the wholesale market is intra-firm. Customer invoices continue to be presented without disaggregated line-item information that would enable consumers to identify the costs incurred at each stage of the supply chain – a level of information disclosure which the Task Force regarded as fundamental to retail competition, but which has never been mandated by Government.59 Possibly the most important lesson from the New Zealand experiment has been the failure of the Task Force’s preferred model of light-handed regulation. Industry self-regulation under information disclosure failed comprehensively over a full decade of attempted implementation. Generators and distributors proved unable to agree on club governance for the grid in 1992–1994. Generators, distributors, retailers, and Transpower were unable to agree 58

Commerce Commission, Decision number 491, www.comcom.govt.nz The New Zealand Commerce Commission, as de facto industry regulator from 2001 to 2003, repeatedly drew attention to this gap in the information disclosure arrangements, with no response from Government. 59

232

Electricity Market Reform

on an industry-wide governance arrangement and rules by 2003, when the Government ran out of patience and established the Electricity Commission.60 After a tentative start, lines companies pushed their margins up from 30% to 70% without triggering any regulatory response from Government. The Commerce Commission (following an inquiry in 2002) retrospectively validated the practice of revaluing the networks’ asset ratebases and hence validated also the radically increased price–cost margins in that sector. The Information Disclosure Regulations introduced in 1994 obliged lines businesses to disclose their financial statements, but New Zealand’s GAAP allowed true rates of return to be hidden in the notes to the accounts, leaving lay members of the public (including, apparently, officials and ministers responsible for oversight of the regulatory regime) in the dark on key issues of pricing and profitability. Looking forward, major new challenges loom on the horizon. New Zealand’s sole large gas field (Maui) is expected to be exhausted by 2007, and only relatively small fields have been located to replace it, raising the possibility that thermal generation will shift to reliance on liquefied natural gas (LNG) or coal. Coal will then be the cheaper thermal option61 unless New Zealand’s compliance with the Kyoto Protocol leads to substantial carbon taxes. In addition, the past 2 years have witnessed large-scale investment in wind farms, which will transform the nature of demands on the grid as wind is matched to (mainly hydro) back-up. Installed wind generation reached 168 MW by the end of 200462 and a further 700 MW of projects are in the planning stage,63 raising the prospect that wind turbines will soon make up over 10% of total generating capacity. Key policy challenges facing New Zealand in the next decade involve dealing with these new issues as well as matters that were ignored or left unresolved in the first round of restructuring. These include the implementation of the Kyoto Protocol to which New Zealand is a party; opening up the demand side of the electricity market to new initiatives such as smallscale distributed generation, time-of-use metering and charging, and net metering of customers with their own generation capability; and breaking the logjam in retail competition. With an electricity regulator at last firmly established, there is an opportunity to make progress on these items of unfinished business.

References Baumol, W.J, Panzar, J.C. and Willig, R.D. (1982). Contestable Markets and the Theory of Industry Structure. Harcourt, Brace and Jovanovich, New York. Bertram, G. (1988). Rents in the New Zealand Energy Sector. Royal Commission on Social Policy, The April Report, Vol. IV. Government Printer, Wellington. Bertram, G. (2003). New Zealand since 1984: elite succession, income distribution, and economic growth in a small trading economy. Geojournal, 59, 93–106. Bertram, G. (2004). What’s wrong with New Zealand’s public benefits test? New Zealand Economic Papers, 38(2), 265–278. Bertram, G. and Terry, S. (2000). Lining Up the Charges: Electricity Line Charges and ODV. Simon Terry Associates, Wellington. 60

Here again there are parallels with the German experience, discussed in Chapter 8. Mighty River Power is in the process of seeking planning permission to convert the mothballed oilfired Marsden B power station to coal. 62 http://www.windenergy.org.nz/FAQ/proj_dom.htm 63 http://www.windenergy.org.nz/ 61

Restructuring of the New Zealand Electricity sector 1984–2005

233

Bertram, G. and Twaddle, D. (2005). Price–cost margins and profit rates in New Zealand electricity distribution networks: the cost of light handed regulation. Journal of Regulatory Economics, 27(3), 281–307. Caygill, D., Wakefield, S. and Kelly, S. (2000). Inquiry into the Electricity Industry: Report to the Minister of Energy. Ministry of Commerce, Wellington, June, http://www.electricityinquiry.govt.nz/reports/ final/index.html Cooper, K. (1995). Valuation techniques and problems: continuing education paper no 439. In proceedings of New Zealand Society of Accountants Infrastructure Assets Forum, Wellington. Easton, B. (1994). Economic and other ideas behind the New Zealand reforms. Oxford Review of Economic Policy, 10(3), 78–94. Electricity Corporation Establishment Board (1987). Electricity Pricing: An Invitation to Contribute. ECNZ, Wellington. Ernst and Whinney (1985). Policy Recommendations for Commercial Efficiency in the New Zealand Electricity Industry. Unpublished consultants’ report, Wellington, August. Ernst and Young. (1994). Rationale for Financial Performance Measures in the Information Disclosure Regime, Including Use of Optimised Deprival Values and Avoidance of Double Counting of Asset Related Expenses: A Report to the Energy Policy Group. Energy and Resources Division, Ministry of Commerce, by Ernst and Young for Briefing ESANZ. Unpublished consultants’ report, Wellington, August. Ernst and Young, Ewbank Preece and State Electricity Commission of Victoria (1991). Valuation of TransPower New Zealand Ltd, Stage 3: Asset Valuation. Wellington, September. Frater, P.R. et al. (1985). Summary Report of Energy Pricing Study and Technical Report of Energy Pricing Study, Business and Economic Research Ltd (BERL), Wellington, June. Hawke, G.R. (1969). Economic decisions and political ossification: The New Zealand retail electricity tariff. In P. Munz (ed.), The Feel of Truth: Essays in New Zealand and Pacific History. Wellington, A.H. & A.W. Reed, pp. 219–233. Hogan, W.W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4(3), 211–242. Hogan, W.W. (1999). The Nodal Zone Debate Revisited, http://ksghome.harvard.edu/⬃whogan/ nezn0227.pdf Jackson, K.E. (1988). Government and enterprise: early days of electricity generation and supply in New Zealand. British Review of New Zealand Studies, 1, 101–121. Jackson, K.E. (1990). Electricity provision and the concept of service in New Zealand. In M. Trédé (ed.), Électricité et Électrification dans le Monde: Actes du Deuxième Colloque International d’Histoire de l’Électricité, Organisé par l’Association pour l’Histoire de l’Électricité en France. Presses Universitaires de France, Paris, July, pp. 411–419. Jarden and Co. (1985). Energy Pricing and Commercialisation. Jarden and Co., Wellington. Martin, J.E. (1998). People, Politics and Power Stations: Electric Power Generation in New Zealand 1880–1998. Electricity Corporation and Department of Internal Affairs, Wellington. Ministry of Energy Financial Objectives and Pricing Review Team (1984). Financial Objectives and Pricing Review for Ministry of Energy (Trading). Wellington. Murray, K. and Stevenson, T. (2004). Analysis of the State of Competition and Investment and Entry Barriers to New Zealand’s Wholesale and Retail Electricity Markets: Report Prepared for the Electricity Commission. LECG and TWSCL, Wellington, August. http://www.electricitycommission.govt.nz/pdfs/opdev/ retail/consultationdocs/pdfsconsultation/pdfscompetition/competition-report.pdf New Zealand Business Round Table (1985). Overview Report: Policy Recommendations for Commercial Efficiency in the New Zealand Electricity Industry. Wellington, August. New Zealand House of Representatives (1992). Report of the Commerce and Marketing Committee: Inquiry into the Proposed Increases of Wholesale and Retail Electricity Prices. Government Print, Wellington, February. New Zealand Treasury (1987). Valuation of Electricity Generation and Transmission Assets. Wellington. Rennie, N. (1988). Power to the People: 100 Years of Public Electricity Supply in New Zealand. Electricity Supply Association of New Zealand, Wellington. Ring, B.J. and Read, E.G. (1993a). Nodal pricing and extensions to the theory. Proceedings of the First ECNZ Optimal Generation Scheduling Workshop, Wellington. Ring, B.J., Drayton, G.R. and Read, E.G. (1993b). Transmission System Pricing Model: Record of Experimental Tests. EMRG Report to Trans Power, Wellington.

234

Electricity Market Reform

Scott, W.D. and Co. (1985). Key Issues Arising from Proposals to Change the Level of Electricity Prices for Bulk Supply in New Zealand. W.D. Scott & Co (NZ) Ltd., Wellington, March. Terry, S., Bertram, G., Dempster, I. and Gale, S. (1992). Hydro New Zealand: Providing for Progressive Pricing of Electricity. Electricity Reform Coalition, Wellington, March. University of Waikato Interfirm Comparison Unit (1985). Comments on Financial Objectives and Pricing Review for the Ministry of Energy. Hamilton, July. Videbeck, S. (2004). Economic Market Segmentation of an Electricity Pool. New Zealand Institute for the Study of Competition and Regulation, Wellington, http://www.nzae.org.nz/conferences/2004/89Videbeck.pdf

Chapter 8 Energy Policy and Investment in the German Power Market G. BRUNEKREEFT1 AND D. BAUKNECHT2 1 Tilburg Law and Economics Center (TILEC), Tilburg University, Tilburg, The Netherlands; 2OekoInstitut – Institute for Applied Ecology, Freiburg, Germany

Summary The authors study the investment incentives of energy policy in Germany and how this affects competition, the environment and supply adequacy. First, after a long period of “self-regulation”, the new Energy Act of 2005 installs a regulator and network regulation. Second, Germany has a strong agenda for the environment. Furthermore, the CO2 emission trading scheme has significant effects. Third, despite international debate, Germany does not have an explicit policy on generation adequacy. The key conclusions are threefold. The initial position of Germany to refrain from regulating network access did not work satisfactorily. The recent creation of regulation can be welcomed and expected to stimulate competition and generation investment. As elsewhere, CO2 permits are allocated free-of-charge, both for existing and new plants. This may be inefficient, but promotes new investment and thus benefits competition and generation adequacy. The data suggests that new generation investment will be required, but also that the market is active. Apart from wind and combined heat and power, coal seems to have a brighter future than sometimes thought. 8.1. Introduction The debate on electric power markets seems to be shifting toward the long-term perspective: does the market provide timely, adequate and efficient investment? Investment (covering both generation and network assets) affects competition, the environment and supply adequacy. In this contribution we analyze and discuss German energy policy with precisely this focus in mind. Evidently, German energy policy does not stand alone, but rather strongly relies on European policy. The policies we examine are threefold. First, the competition and network regulation of the electric power market, which changed direction in 2005. Whereas Germany initially took an exceptional position within Europe, it is now more in line with neighboring countries. Second, environmental policy and in particular the start of the CO2 emission trading scheme deserves attention. Germany has a strong agenda for promoting environmentally friendly technologies and energy efficiency. The question is how much scope remains for pursuing these objectives. Third, it is noteworthy that there is no policy on generation adequacy; in the light of experience and growing concern in other countries the question is whether this is justifiable. 235

236

Electricity Market Reform

The capacity margins are now declining but still comfortable, stemming as they do from excess generation capacity before liberalization. Investment levels have been picking up again in the last 2 years after a serious decline in the previous 5 years which followed on from an all-time high as a result of modernizing the former east following reunification in 1990. This may reflect better competitive opportunities indicated by higher wholesale prices. More worrying is that generation assets are old and need to be replaced, while it is unclear what should replace these. The share of coal and lignite in the generation mix is already large. The CO2 reductions required by the Kyoto protocol raise doubt about the future of coal. The share of nuclear is already 30% and is being phased out, while new build lacks support. Support for renewables (RES) is strong but it is unclear whether this has sufficient scope. Our key conclusions are also threefold. The initial exceptional position of Germany to refrain from (ex-ante, sector specific) regulating network access did not work satisfactorily. Clearly, the institutions were not in equilibrium. The recent creation of a regulator (the Bundesnetzagentur, BNA) and regulation (with the new Energy Act of July 13, 2005) can only be welcomed. For a variety of reasons, network regulation must be expected to promote competition and thereby stimulate new investment by newcomers. We note that the wholesale margin increases. As elsewhere, CO2 permits are allocated free-of-charge (instead of auctioned), both for existing plants as well as for new investment. Evidently this has a political background and cannot be supported on economic grounds. Still, a free-of-charge allocation does promote new investment, and thus it may be inefficient, but benefits competition and generation adequacy. A curiosity is the so-called transfer rule which grants new (CO2 poor) plant the number CO2 permits of the CO2 rich plant it replaces. This is good for the environment, but sets new entrants at a serious disadvantage. With respect to generation adequacy (taking into account the points made above), we note that data suggests that new investment is required, but also that (as elsewhere) there appears to be a lot of new construction plans. Apart from wind and combined heat and power (CHP), new coal seems to have, perhaps surprisingly, a brighter future than sometimes thought. The organization of this contribution is as follows. Section 8.2 provides an overview of the recent history and the current state of the German electricity supply industry (ESI). Section 8.3 gives an in-depth examination of various energy policies distinguishing between the Energy Act, environment policies and generation adequacy. Section 8.3 includes analysis of the investment effects. Section 8.4 is the conclusion.

8.2. The German Electricity Supply Industry 8.2.1. How the sector is structured With a population of 82 million, Germany has the largest power market in Europe. Total net electricity consumption is around 500 TWh/year; installed gross capacity is around 140 GW including more than 15 GW wind capacity – more than any other country in absolute terms. Peak load in 2004 was at 77.2 GW. Total gross revenues in the sector are roughly €60 billion/ year and investment amounts to around €4 billion/year. Germany’s transmission network is integrated with that of nine neighboring countries. Imports and exports are more or less balanced, with a total of 44 TWh imports and 51 TWh exports in 2004. The exchange with France, the Netherlands and the Austrian and Swiss hydro systems is especially significant (see Table 8.1).

237

Energy Policy and Investment in the German Power Market Table 8.1. German power imports and exports (in TWh). 2004

Austria Switzerland France Luxembourg The Netherlands Denmark Czech Republic Poland Sweden Total

2003

Imports

Exports

Imports

Exports

4.4 2.8 15.5 0.8 0.6 5.3 13.1 0.4 1.3 44.2

8.9 11.8 0.4 4.9 17.3 3.4 0.1 3.2 1.5 51.5

3.3 3.1 20.2 0.8 0.6 4.0 12.8 0.3 0.6 45.7

9.9 13.2 0.2 5.0 15.0 5.4 0.1 2.8 2.2 53.8

Source: VDN Berlin. 700 600 Other 500

Gas

TWh

Oil 400 300

Coal Lignite Nuclear

200

Hydro

100 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Fig. 8.1. Generation mix. Source: Brunekreeft and Twelemann (2005).

The main energy source for power generation is coal, which accounts for around 50% of electricity production, with hard coal and lignite each accounting for about half of this (Fig. 8.1). Germany has substantial domestic coal reserves. Yet German hard coal is about three to four times as expensive as imported coal and relies on state subsidies. Since the number of miners has become too small to be a serious interest group1 and the hard coal subsidy is a state aid and is for this reason not favored by the European Union (EU) commission,2 the subsidy is gradually reduced. This will not influence the generation mix or investment decisions, which are based on the price of imported coal, but will reduce domestic coal consumption and increase coal imports. Lignite is the only other major domestic energy source in Germany. As it can be accessed through open-cast mines, it is relatively cheap and does not require state subsidies.3 The 1

In 2003, there were just over 50,000 people working in hard coal mining and processing and around 15,000 in lignite. The number of employees is bound to decrease further. 2 Cf. EU Regulation, No. 1407/2002; July 23, 2002.

238

Electricity Market Reform

downside is that open-cast mines consume vast chunks of land, leading to significant public opposition. The RWE utility in the West and Vattenfall utility in the East are the main lignite producers and generators. While there was a major overhaul of plants in Eastern Germany after reunification, RWE operates a much older fleet of lignite plants, with quite a few plants approaching their 50th anniversary. Nuclear plants generate about one-third of the total power production. However, the red–green government coalition agreed on a nuclear phase-out program which was laid down in the 2001 amendment of the nuclear law. The agreement stipulates a generation limit based on a 32-year plant operation, which means that nuclear generation will be phased out at around 2020 according to this plan. However, companies have the option to shift generation allowances between plants to increase the output in more efficient plants. In mid-2005, only two plants: Stade and Obrigheim, were closed. The conservative party has announced that it wants to do away with this agreement and extend the plants’ lifetime. However, with the likely new conservative-socialist coalition it is unclear what will happen. There has been no “dash for gas” yet, with gas still accounting for only around 10% of power generation. There are only a few combined cycle gas turbines (CCGT) plants. With only minor domestic gas supplies and a high dependency on gas imports, mainly from Russia, the Netherlands and Norway, there is some concern that an increasing share of gas may undermine supply security. Despite this, generation from gas is forecasted to increase in most scenarios. For instance, a report for the Ministry of Economics (BMWA, 2005, p. 33) projects a share of gas of about 33% by 2030. The remaining power is generated from hydro plants and a rapidly increasing share of “new” RES, especially wind. The above-mentioned report for the Ministry of Economics also projects a share of about 33% for RES in 2030. Demand growth is relatively low at around 1% per year and is expected to remain at this level.4 CHP has a production share of about 10%,5 of which 60% is gas fueled and 40% hard coal (and lignite) fueled (Fig. 8.2). There are no official statistics on distributed generation, but the generation share is reported to be at around 18% (Wade, 2005). 160 CHP Power only

140 120 TWh

100 80 60 40 20 0

Gas

Hard coal

Lignite

Others

Fig. 8.2. CHP generation. Not including industrial plants. Source: Destatis. 3

There are no official and explicit subsides, yet it can be argued that there is hidden financial support (Wuppertal Institut, 2004). 4 Source: VDEW (www.strom.de). 5 This does not include industrial plants, which are often CHP plants.

Energy Policy and Investment in the German Power Market

239

Implementing the EU Directive of 1996, the German ESI was liberalized in 1998 with the Energy Act of 1998. The sector was never institutionally monopolized (like for instance the UK). Instead competition in a relatively unconcentrated and fragmented industry was excluded by cartels agreements, which were stabilized by legally enforced demarcation contracts. The main step of liberalization was to invalidate these cartel agreements after which the ESI fell under the authority of the Cartel Office and the Competition Act. The industry structure, which was artificially stabilized by the demarcation contracts, strongly reflected historical institutional lines; roughly speaking, generation and transmission reflected the position of the states within the federal structure of Germany, while distribution and retail reflected the strong position of the communities. Not surprisingly and as we have seen elsewhere, competitive pressure and commercial interests enforced significant changes in the industry structure after liberalization, most notably toward more concentration. Vertically, the ESI was strongly integrated between networks and commercial businesses and if anything this has increased since liberalization; it is thus remarkable that the Energy Act of 1998 only required minimal vertical unbundling requirements. We will discuss this in more detail below. The German ESI is strongly vertically integrated. There are basically two blocks, which is depicted in Figure 8.4. On the one hand, four predominantly privately owned big utilities own and operate the high-voltage transmission grids (plus the interconnectors) and most of the power plants (usually in their own control area). These firms are both transmission system operators (TSOs) and dominant generators. They operate the balancing market in their own control area. There is an ongoing discussion about separating these balancing markets and merging them into a national balancing market. Together these four companies own about 90% of total generation capacity (Table 8.2). Moreover, they also have majority shares in many distribution networks and retail activities. These four utilities are RWE, E.On, EnBW and Vattenfall Europe. On the other hand, a vast number of predominantly municipality-owned firms (Stadtwerke) own and operate the distribution networks and, as end user switching away from the incumbent retailer has been low, mostly the retail activities in the subsequent host areas.6 As we will argue below, and as also noted by Haas et al. (2005) the high degree of vertical integration led to cross-subsidization between networks and generation, stifling competition. Most of the municipal utilities were considered to be too small and expected to disappear quickly after liberalization. However, most of them have done much better than expected, setting up various alliances to defend their market position and to be able to take part in wholesale trading and realize economies of scale, for example, in billing. The vertically integrated “Stadtwerke” also responded to market opening by lowering the retail margins (being the difference between the end user price and network charge plus wholesale price), making life for new third-party retailers difficult. Cumulative domestic switching rates are reported to be 5%. Mergers and acquisitions have increased concentration in generation since the beginning of liberalization (see Table 8.2). Around 2000, two big mergers, creating the current firms RWE and E.On, pushed the Hirschman Herfindahl Index (HHI) to more than 2500.7 6

The exact number is unclear. VDN, the association of network operators, has 390 members. Also, VDN lists in its publication of network charges 700 different networks some of which fall under the same holding though. VDEW, another industry association mentions 900 firms. 7 In European merger control a post-merger HHI of 2000 and in the USA of 1800 are crucial thresholds. Note however that these are very rough indications, which neglect many details.

240

Electricity Market Reform

Berlin 7 2 4 Dortmund

3

Bayreuth 1

Stuttgart 3

1 EnBW Transportnetze AG 2 E.ON Netz GmbH 3 RWE Transportnetz Strom GmbH 4 Vattenfall Europe Transmission GmbH Fig. 8.3. German transmission system operators.

8.2.2. The institutional steps Germany implemented the 1996 EU Electricity Directive (EU E-Directive) with the Energy Act of 1998, of which three aspects stood out. First, full market opening. For generation this is unsurprising, but 100% end user eligibility from the start was exceptional in 1998. Even

241

Energy Policy and Investment in the German Power Market

Generation

4 Verbundunternehmen (VUs)

Transmission

~900 communal distributors

Distribution

Distribution

Retail

Retail

Fig. 8.4. A stylized representation of the ESI in Germany. Table 8.2. Market shares in generation (percentages of output). 1994

VEBA VIAG RWE VEW EVS Badenwerk HEW BEWAG VEAG Other Total HHI

} E.ON } RWE } EnBW } V’FALL

2000

A

B

Pre A

Pre B

Post

16.92 11.23 31.38 7.24 4.89 4.91 3.55 2.87 – 17.00 100 1807

13.96 8.27 28.42 6.65 4.30 4.32 2.96 2.28 11.84 17.00 100 1595

21.36 12.55 31.53 8.84

18.77 9.97 28.94 8.33

} 28.74

} 9.64

} 8.60

} 8.60

3.09 2.65 – 10.35 100 1903

2.57 2.13 10.33 10.35 100 1658

} 37.27

} 15.03 10.35 100 2622

HHI: Herfindahl Hirschman Index. Notes: The shares have been corrected for participation rates. Pre means Pre-merger and Post means Post-merger. V’Fall is Vattenfall. Source: Brunekreeft and Twelemann (2005, p. 103).

early 2005, only nine European member states had full retail market opening. And although the EU E-Directive 2003 aims at full retail market opening by 2007, we expect that there will be a debate on whether this will be pursued. Full retail competition in Germany worked well technically, but competition developed only slowly for domestic and small commercial end users. Second, whilst the degree of vertical integration of monopolistic networks and commercial businesses is high and increasing, the rules on unbundling were weak and were not enforced. Third, being the exception within Europe, Germany opted for negotiated Third Party Access (TPA), instead of regulated TPA.8 Negotiated TPA implied that, despite the monopolistic networks, the sector was left without sector-specific regulation and regulator. The government trusted the ESI to resolve 8

Cf. Haas et al. (2006) for a European overview. Moreover, Brunekreeft (2003, pp. 208 ff.) contemplates on possible explanations for this exceptional position; it is rather likely that the reunification in 1990 contributes to an explanation.

242

Electricity Market Reform

network access and network charges by voluntary negotiations controlled by the Cartel Office. Network access had to be arranged collectively in the so-called association agreements (VV). Initially, these arranged the (technical and administrative) rules but not the price of network access. Later, a set of accounting principles to calculate the network access charges was added to the VV. At no stage though was the precise level of network charges agreed upon or laid down. These were the sole responsibility of the individual network owners. Both the network access and the network charges were controlled by the Cartel Office. To facilitate this task, the Competition Act was strengthened with an essential facilities doctrine in 1998, which requires that access to the network should be provided on nondiscriminatory terms and at a fair and reasonable charge. This one clause in the Competition Act was the main regulatory instrument.9 Control was not strong and network charges were (and in fact still are) persistently high. The Cartel Office faced a number of problems (Bundeskartellamt, 2001). First, it is allowed to act only after a justified suspicion of abuse; hence, it can act only ex-post. Second, with up to 900 networks to be controlled, the Cartel Office was seriously understaffed for this task. Third, many of communal and regional networks enjoy political support from the states and communities. Fourth, the Competition Act is well suited to address discriminatory behavior; the more persistent problem, however, turned out to be the high level of the network charges, which is difficult to address with the Competition Act. Lastly, accounting according to the association agreement received legal validity, which in practice weakened the position of the Cartel Office. After a series of events and reports, the so-called Monitoring Report of the Ministry of Economics paved the way to stronger regulation in 2003. Parallel to this, the German government gave up its resistance in Brussels and the European Commission seized the opportunity to remove negotiated TPA from the directive. Hence, the new EU E-Directive 2003 exclusively allows regulated TPA. This is also the key development which led to the new Energy Act which entered into force on July 13, 2005. 8.2.3. Past, present and future Figure 8.5 plots an interesting development. Shortly after liberalization end user prices (net of taxes) fell strongly, but started rising again shortly afterwards and quite steeply since a year or so.10 The figure depicts the representative domestic user of Eurostat and relies on Eurostat data. This implies that it only captures the non-switching part of the market, which are still under the old tariff regime. It does not capture the prices of the competitors and thus the prices for switching end users. As shown elsewhere (Brunekreeft, 2003, p. 220), the best-practice alternative offers undercut the incumbent price severely at first, but then started to increase and converge. Meanwhile the difference is small. Further we should remark that the domestic market excludes the industrial users. The pattern for industrial prices is the same but far more dramatic. Industrial prices have fallen severely but are now increasing steeply as well. A steady increase of the electricity tax is an important contributor to higher prices. It has been raised gradually over the last 6 years to up to €2.05 c/kWh, which makes up somewhat

9

This is not unlike the situation in the USA under the Energy Policy Act 1992. Order 888 of 1996 made a strong move toward regulation of network charges (cf. Joskow, 2005a). 10 Cf. also Growitsch and Müsgens (2005) for more details.

243

Energy Policy and Investment in the German Power Market 17

c/kWh

16 15 14 13 12

Germany w/o tax EU 15 w/o tax

11

Germany with E-tax January 2005

January 2004

January 2003

January 2002

January 2001

January 2000

January 1999

January 1998

January 1997

January 1996

January 1995

January 1994

January 1993

January 1992

January 1991

10

ct/kWh

Fig. 8.5. Residential end user prices in Germany and Europe-15. Source: Eurostat data, various years. Note: This is for an average domestic end user; eurocent/kWh. Nominal prices. 8 7 6 5 4 3 2 1 0

VAT (16%) Concession fee RES CHP E-tax 1998

1999

2000

2001

2002

2003

2004

Fig. 8.6. Taxes in the German electricity price. Source: VDEW.

more than 11% of the total price. Further substantial parts are the communal concession fee and the federal value added tax; however, these are stable and do not explain the increase. While these taxes are substantial, they are unambiguous. More ambiguous are two levies induced by support for CHP11 and RES.12 These are not strictly speaking taxes, but the costs of CHP and RES are socialized over network users and electricity-consumers, respectively. It is clear from Figure 8.6 that whereas these costs are increasing steeply in relative terms, they are unsubstantial in absolute terms. Yet, the industry sometimes justifies the price increases, especially the increased network charges, with these “tax” increases. The development of the network charges is ambiguous. Up to mid-2005 the network charges were unregulated. At best, one could argue that rather loose self-regulation was enforced by either some threat of ex-post control by the Cartel Office, or by the threat of a change toward ex-ante regulation, which indeed happened in mid-2005. In the course of

11 12

Combined heat and power. Renewable energies.

244

Electricity Market Reform

65.00 55.00

/MWh

45.00 35.00 25.00

Peak Peak minus CO2 CO2 price CO2 coal markup

15.00

September 2005

July 2005

August 2005

May 2005

June 2005

April 2005

March 2005

January 2005

February 2005

December 2004

October 2004

November 2004

September 2004

July 2004

August 2004

May 2004

June 2004

April 2004

March 2004

5.00

Fig. 8.7. Wholesale prices and CO2 mark-up at EEX. Source: EEX.

self-regulation the association of network operators (VDN) started to publish a standard format of a sample of network charges for different voltage levels twice a year. Examination of the LV network charges reveals that they were high in international comparison (cf. e.g. EC, 2005) and high relative to end user prices (cf. Brunekreeft, 2002). However, they have been stable since at least 2002. Only the HV level has seen an increase of about 10%, which is unsubstantial in absolute terms. According to the network operators, this increase is just the cost-pass-through of higher balancing costs due to an increase in intermittent RES generation. A study of the association of industrial users, VIK13 suggests a recent increase in the network charges for industrial customers for selected networks. It is unclear whether the changes are representative for the sector. Growitsch and Wein (2004) calculate a reduction of the spread in network charges among various operators as a result of the introduction of the self-regulation in VVII⫹.14 This suggests that the increases by some are leveled out by decreases by others. All things considered, we should conclude that the recent increase in end user prices cannot be explained by changes in network charges. Much of the recent price increase can be explained by the wholesale price development. We note that more than 90% is traded “over the counter”, for which we do not know the prices. However, we assume that the spot price at the European Energy Exchange (EEX) in Leipzig is a sufficiently good indicator for contract prices. Figure 8.7 gives the EEX price development and clearly shows that the price is rising. The wholesale prices used to be very low. In fact, after liberalization the prices went down to almost short-run marginal costs and could not recover total costs. This is changing. With well over €35/MWh, it is believed that full cost recovery has been restored. Why the recent increase? There are three plausible explanations.

13 14

Cf. www.vik-online.de, April 28, 2005. VV abbreviates the German word Verbaendevereinbarung.

245

Energy Policy and Investment in the German Power Market 60

Demand

/MWh

50

Merit order stylized CO2: 20 /tonne

40

Price

30 Gas

20 10 0

Coal Hydro Nuclear 0

20

Lignite

GW

40

60

80

100

Fig. 8.8. The merit order and subsequent marginal prices in Germany with a CO2 price of €20/tonne CO2. Source: Bremer Energie Institut (this was handed to the authors by Wolfgang Pfaffenberger). 30.00 25.00

/MWh

20.00 15.00 10.00

Gas (th) Coal (th) Gas (el) Coal (el)

5.00 0.00 1998

1999

2000

2001

2002

2003

2004

2005

Fig. 8.9. Price development of gas and coal. Source: BAFA (efficiency gas: 59% and coal: 45%).

First, with the start of emission trading (see below) the electricity wholesale price now includes a CO2 mark-up. With a CO2 price of €20/tonne CO2 and an emission factor for a coal plant of 0.75 t/MWh, this amounts to a mark-up of €15/MWh on the wholesale electricity price. If gas is marginal, then the mark-up is about €7/MWh because the emission factor of gas plants is about 0.35. If generation is reasonably competitive then the CO2 markup would by and large be passed through into the EEX price. In either case, the effect of CO2 prices is substantial, as shown for real values in Figure 8.8. More importantly, Figure 8.7 suggests that if the CO2 mark-up is subtracted the net wholesale price fluctuates around €35/MWh, which corresponds by and large to the prices in 2004. Second, the fuel prices have increased as depicted in Figure 8.9. For Germany the increase in gas and oil prices has less influence on wholesale prices because gas is not often marginal, but the world coal price has increased recently as well. Third, for different reasons, the price increase may reflect a stronger control of the producers on the market, which at the very least seems to ease competitive pressure. Using an elaborate electricity market model, Müsgens (2004) makes an in-depth examination of costs, bidding and prices in the period from 2000 to mid-2003. He finds a structural break in early

246

Electricity Market Reform

2002. Before that prices were closely in line with system marginal costs, while from then on prices started to diverge from system marginal costs. Notably, the divergence is mostly in high demand periods. It does not follow though that these prices are excessive; as already noted, the early prices were too low to recover costs, whereas after early 2002, they might just recover full cost. Why would prices start to diverge from system marginal costs? Arguably, installed capacity in Germany was excessive, which was likely to increase competitive pressure and suppressed prices down to short-run marginal costs. Around 2000, the two big players, E.On and RWE announced that they would shut down some 10 GW of plant capacity because prices were too low; part of this was decommissioned and part mothballed. In addition, to other reasons a decline of excess capacity and the resulting relative scarcity are likely to have given the firms some grip on the prices.15 Alternatively, the higher prices might simply reflect emerging scarcity and actually signal that new investment is needed. Further, as shown in Table 8.2, concentration has increased around 2000 with the RWE–VEW (now RWE) and VEBA–VIAG (now E.On) mergers. With the HHI increasing from 1700 to 2600, theoretical insight and experience abroad would suggest a potential weakening of competitive pressure, as also suggested by Haas et al. (2005). Cross-border trade will certainly increase competitive pressure, but this is still limited. Table 8.1 shows that Germany is a net exporter, mainly because the prices in the Netherlands are higher than in Germany. Furthermore, cross-border capacity is only some 14% of total installed capacity. We would like to stress though that current wholesale prices (less of the CO2 mark up) do not seem to establish an abuse of market power.16 Lastly, in 2001 a report of the Cartel Office made clear that network charges were high (Bundeskartellamt, 2001). As a result the industry attempted to strengthen industrial selfregulation and started to publish the network charges systematically and in a comparable way. Pressure to regulate network access and network charges started to increase. The EU E-Directive 2003 removed negotiated TPA altogether and required regulation, which has now taken shape in Germany (see below). As argued extensively in Brunekreeft (2002, 2004), given vertical integration and the lack of effective regulation of network revenues, the rational strategy was to concentrate on the network while keeping the margins in the competitive businesses low and thereby retaining market shares. As regulation of the network takes shape we would expect the reverse, that is, an increase in both the wholesale and the retail margins. The implications for new entry and investment follow swiftly. After the German market was liberalized in 1998, foreign companies showed especial interest in entering the German market. Although the German generation market was by far not as attractive as, for example, the Spanish or Italien markets in terms of wholesale price, the need for additional capacity or expected demand growth, many companies considered it strategically important to be present in the largest European electricity market. The large number of small- and medium-sized utilities provided good take-over candidates. It took only a few years for this excitement to die down. This was partly because many new players like Enron or Dynegy either completely disappeared or were on the brink of disappearing. More importantly, the German market turned out to be hostile toward new

15

In formal terms, mothballing capacity can be interpreted as a credible commitment not to use this capacity and it can thereby ease competitive pressure. 16 This leaves the question open what exactly is market power and how prices above marginal costs can be stable in a competitive environment.

Energy Policy and Investment in the German Power Market

247

entry for a number of reasons. First, it was difficult to get hold of new plant sites. Second, as indicated above, wholesale prices were unattractively low. Third, in the first few years after liberalization the arrangements on network access were biased against third parties. Fourth, there have been persistent complaints about discrimination of third parties. Fifth, as far as new entry from gas-fueled CCGT is concerned, the big electricity–gas merger E.On-Ruhrgas (controversially approved in 2002) did not improve competitive conditions. Sixth, a gas tax represented a significant entry barrier for new gas plants. The gas tax increased the costs of a new CCGT plant by ca. €3/MWh or 10%. There were limited exemptions from the tax for new CCGT plants with an efficiency of more than 57.5%. Following a new European directive on energy taxation,17 this tax was abolished in 2005. Many new plant projects were either given up or had to look for new investors who were willing to keep these projects alive and wait for better times. New firms entered the market mainly through acquisitions; for instance EdF bought a minority stake in EnBW and Vattenfall took over Bewag, HEW and VEAG. This did not always improve competition, but rather increased the concentration in the wholesale market. While Germany has not seen many new plant projects that actually came on-line since market opening, investment activity is now picking up again as we will indicate in Section 8.3.3. The wider wholesale margin is definitely helpful in this respect.

8.3. Energy Policies and the Investment Effect Before discussing energy policies in Germany in detail, Table 8.3 gives an overview of the main acts and events as they affect the ESI. 8.3.1. The Energy Act 2005, regulation and the regulator BNA 8.3.1.1. The Energy Act 2005 As explained above, for a variety of reasons, the Energy Act 1998 was replaced by the Energy Act 2005. The key elements in the new Energy Act 2005 are the following. First, the rather artificial hybrid approach of ex-ante approval of the methodology and ex-post control of the level did not survive the debate, and a clear step has been taken toward ex-ante regulation of the network charges. Second, although starting off with a cost-based approach, it is an explicit intention of the authorities to switch to incentive-based regulation. Third, there will be a sector-specific regulator: the BNA. Fourth, the rules on unbundling are strengthened but they still only minimally fulfil the directive’s requirements. We will discuss these points in detail in this section.18 Ex-ante regulation of the network charges. Art. 23(2) of the EU E-Directive 2003 requires “fixing or approving, prior to their entry into force, at least the methodologies used to calculate” the network charges. The precise phrasing reflects the German wish to stick to ex-post control of the level of the charges. In fact, this type of ex-ante/ex-post hybrid regulation has been practiced in, for instance, Sweden (with mixed success) and Finland (where it worked well). After a long debate, it has been decided in Germany that the by-pass in the directive

17

Directive 2003/96/EG. Haas et al. (2005) are somewhat sceptical about the new Energy Act and point out that the legislator might have taken the opportunity to put in place a more pro-competitive market design. 18

248

Electricity Market Reform

Table 8.3. Major events in the German electricity supply industry. Date

Event

Comments

April 1998

Amendment Energy Act, start of liberalization

100% market opening in generation and supply.

May 1998

First Association Agreement VVI

Self-regulation of networks.

April 1999

Introduction of the electricity tax

Starting with 1.02 c/kWh and gradually increased to 2.05 c/kWh.

December 1999

Second Association Agreement VVII

New retailers (e.g. Yello) enter the market; prices drop severely and surprisingly, but only short lived.

April 2000

Renewable Energy Act

Fixed feed-in tariffs for renewables.

August 2000

EEX in Frankfurt

December 2001

New Nuclear Act

Stipulates phase out of nuclear plants in Germany, prohibits new nuclear plants.

April 2001

Report of the federal Cartel Office

Indicates that network charges are high and difficult to control by Cartel Office by Competition Act.

December 2001

Third Association Agreement VVII⫹

Stronger emphasis on industrial self-regulation.

2000/2001

Mergers

RWE and VEW: RWE VIAG and VEBA: E.On

2002

Merger

E.On and Ruhrgas

December 2003

Monitoring report by the Ministry of Economics

This report confirms “officially” that negotiated TPA in the ESI and GSI failed. It paves the way to regulated TPA.

January 2005

Emission trading starts

July 2005

Amendment of Energy Act, Regulatory Authority (BNA) takes over electricity network regulation

Implementing EU Directive, ending 7 years of self-regulation of the network.

GSI: gas supply industry.

be ignored and the ex-ante regulation of the level of the access charges be applied. Thereby the regulation of the network access charges is finally as it should be.19 There has been some debate about controlling price increases only. This would imply that all current levels would be beyond the authority of the regulator. Since some network operators have increased their charges quite significantly over the last year (i.e. before regulation would take effect), this restriction in regulatory authority was unacceptable and was overthrown. The regulator has now been authorized to look back and control the recent price increases. Cost-based versus incentive-based regulation. The main debate has been on the type of regulation. The formal current state is that regulation is cost based (para. 21), which reflects 19

Increases of domestic end user prices required approval of state authorities relying on a federal decree. The enforcement of this decree has always been questioned. In any case, as a result of the new regulation of the network charges, this decree on end user prices will expire in mid-2007.

Energy Policy and Investment in the German Power Market

249

business as usual. Previously, the “self-regulation” followed the accounting principles laid down in the association agreement. This is nothing else than a rate-of-return regulation, with the difference that it will now be enforced and resulting charges will have to approved before they enter into force.20 The legislator explicitly allows the option to switch to incentive-based regulation (para. 21a), which can be a price-cap or revenue-cap regulation. The regulator has been given the task to develop an incentive-regulation mechanism. However, whether, how and when this will be implemented is to be determined by the government in an ordinance (i.e. not by the regulator).21 The choice between cost-based and incentive-based regulation deserves attention. Incentive-based regulation aims at improving the incentives of the regulated firm to produce efficiently (i.e. cut costs). The means to do so is to allow the firm to keep the profits resulting from efficiency improvements. Not having to lower the (ex-ante allowed) prices after lowering costs for some predetermined period is the incentive. The German Energy Act explicitly mentions incentive-based regulation, but as has been pointed out by Joskow (1989, 2005b), it is not so clear what this means and how incentive-based differs from cost-based regulation.22 Three points seem important. The name rate of return regulation (being a typical form of cost-based regulation) suggests that prices should always be adjusted to costs so as to allow a reasonable rate of return. In practice this is not the case as rate-of-return regulation typically also fixes (weighted average) prices for some period of time: the regulatory lag. During this period prices can deviate from the “fair” rate of return. The difference in emphasis is that under typical cost-based regulation the regulatory lag is endogenous and relatively short. As Joskow (1974) explains well, typically in the USA, (weighted average) prices remained fixed until either the firm or the regulator requested a rate hearing. An important innovation of the incentive-based regulation first introduced by Littlechild in 1983 was to make the regulatory lag explicit and exogenous as a closed regulatory contract (cf. Beesley and Littlechild, 1989). A second point is that a cost-based approach typically adds a mark-up to the firm’s own costs. This is fair and reasonable but does little for the incentives to keep costs low. An incentive-based approach steps away from this and tries to avoid the use of the firm’s own costs as the benchmark. Instead it might use an industry benchmark. This retains the incentives but may lead to unreasonable results. These are theoretical polar cases; in practice the difference is blurred. Typically, with cost-based regulation, the regulator will look at whether the underlying costs are reasonable and thereby use comparators. Also, in incentive-based regulation any regulator will always check whether the outcome for an individual firm is reasonable.

20

The interested reader may refer to Brunekreeft and Twelemann (2005, p. 109) for details. The new Energy Act is a step away from the controversial accounting of replacement value. Presumably the practical background is that replacement value can lead to a high RAB whereas the network may have completely depreciated. 21 It may be noted as an aside that the policy uncertainty is striking: it is, at best, likely that regulation will be incentive based in the future, but we do not know when or what it will look like. This may be compared to Norwegian legislation where the switch to incentive-based regulation in 1997 was laid down in 1991 in the law. 22 To be precise, incentive-based regulation is the overarching term of which pure cost-pass-through and pure price-cap are the polar cases (cf. Joskow, 2005b). So, it is not really appropriate to contrast cost based with incentive based. We will assume that in the Energy Act incentive-regulation means a move toward price or revenue capping.

250

Electricity Market Reform

A third point is that pure cost-based or pure incentive-based mechanisms only exist in textbooks. In practice details matter and we find all kinds of adjustments and modifications and we see that the polar cases converge. Two examples are important. Rate-of-return regulation can be modified by a use-and-useful clause (cf. Joskow, 1989), which basically says that the investment costs may only be passed through in the rate base if the investment is used and useful, on which the regulator decides. Clearly this steps away from pure costbased regulation. Incentive-based regulation can be adjusted by profit-sharing rules, basically saying that if under the incentives-based constraint profits get either too large or too small, prices can (should) be adjusted. This clearly adds a cost-based element. Illustratively, Grout and Zalewska (2003) define a profit-sharing rule as a weighted average of the outcomes under cost-based and incentive-based regulation.23 The Energy Act (para. 21a.2) highlights the ex-ante determination of the average revenue cap as the decisive point. The control period will be between 2 and 5 years. Furthermore, relative efficiency will be determined by benchmarking with comparable firms.24 It should be noted though that this also holds for the current cost-based approach, where the reasonableness of the firm’s own costs can be checked by comparing with other firms. Moreover, only the costs components, which are under control of the firms will be subject to efficiency incentives. Although without explicit details, the Act touches upon the following aspects. First, presumably the price-cap regulation will be tariff basket, capping the weighted average price of a basket of products and leaving individual prices to the firms.25 Second, the regulation will explicitly be quality adjusted, presumably with a penalty-and-reward system. Third, it seems unlikely that there will be a yardstick; the Xi will be firm individual or firms will be collected in comparable groups. The discussion on X versus Xi is non-trivial in the face of up to 900 networks. Faced with so many firms, a yardstick X for all is very attractive but seems unreasonable. Bundesnetzagentur. The EU E-Directive 2003 requires with art. 23(1) regulatory authorities, “wholly independent from the interest of the electricity industry.” This excludes industrial self-regulation as it was practiced in the German ESI, especially by means of the VVII⫹. The new Energy Act creates the sector-specific regulator BNA, which will include the regulator for gas, telecommunications and postal services and which will also cover railways. Authority has been split though. The federal regulator BNA is responsible for all network operators with more than 100,000 customers (and for network owners with less than 100,000 customers that operate in more than one state). The states are in charge of regulating smaller network operators. However, if desired, the states can hand over the regulation to the federal BNA.26 This follows art. 15(2) in the EU E-Directive 2003, which exempts network operators with less than 100,000 customers from unbundling rules (except separate accounts). At least 500 networks are the responsibility of the states.27 As the communal lobby is very strong, and states and municipalities are the main stakeholders in the distribution network operators (DNOs), we may expect that state regulation of the DNOs will be weaker than federal regulation. 23

The distribution price control 2005–2010 in the UK provides interesting examples. There is some discussion to apply a virtual network approach as in Sweden to pre-select some very highly priced networks. 25 This stands out against the regulation of telecommunications, which has a stronger leg in the regulation of individual prices. 26 In Summer 2005, most states have decided to keep a state regulator. 27 However, the aggregate market covered by these firms will be small.

24

Energy Policy and Investment in the German Power Market

251

It seems that all the regulators will have to follow the same federal ordinance concerning the choice of regulation. This is a missed opportunity. As pointed out the idea is to switch to incentive-based regulation. One of the problems is how to manage the regulation of 900, mostly very small DNOs. Exactly this problem could be by-passed by applying different types of regulation: a strict incentive-based regulation at federal level and a “loose” costplus approach at state level for many small utilities. If all the small DNOs are also regulated by the same type of incentive regulation, it is unclear what is gained with splitting up the authority, while it opens the door for regulatory capture at state level. Unbundling. The unbundling requirements correspond to the EU Directive. Hence, the Energy Act requires legal (and functional and management) separation of TSO and DSO (with the art. 15(2) exemptions as mentioned above), confidentiality of information and accounting separation. This has by and large been implemented. The more urgent point is whether it will be enforced and controlled. As the regulator will pick up its task, we are confident that this will indeed be serious and that firewalls will start to be pressing. The more interesting question is whether ownership unbundling has any prospect. This question is aimed at the TSO in first instance. The legal problem is that the four TSOs are largely in private hands and that ownership unbundling is expropriation and violates the constitution. However, there are signs that ownership unbundling may re-emerge as an issue. First, there is some debate to split off the system operators (SO) from the rest of the firm and thus leave the transmission ownership (TO) to the current owners. This would also allow the creation of both one national SO and one national balancing market. The SO has no assets and this approach would therefore most likely not to be regarded as expropriation. Alternatively, all current firms could have the national SO in collective ownership. Second, experience in the UK, for instance, suggests that very strict firewalls can make “voluntary” unbundling an attractive option for the companies. Typically, this requires very strict monitoring by the “watchdog” and hence depends a great deal on the BNA. 8.3.1.2. The institutional disequilibrium Why did the government decide not to regulate from the beginning of liberalization? Recall that the German telecommunication sector does have sector-specific regulation by a regulator. Although speculative, four arguments are apparent. First, the legislator may not have been completely benevolent. The sector’s influence on politics is considerable. Second, the energy sector (gas and electricity) is considered to be strategic. Faced with counterparties like Gazprom, the government hesitates to fragment the industry too heavily and tries to balance between different goals (in this case, in particular between competition and countervailing power). Third, there has also been the desire to create and support “national champions” able to compete on a European scale. Fourth, after reunification the firms from the West committed to investing heavily in the former East in order to modernize both plants and networks. Oddly, this did not result in stranded-costs claims when liberalization started, unless we should interpret the lack of regulation as such. In any case, only 7 years after liberalization, the institutional framework was adjusted to adopt ex-ante regulation of the network charges. Hence, we may conclude that the framework was not in equilibrium and that something went wrong.28 Changes in the ESI are at least partly a spin-off of the gas supply industry (GSI). The GSI as well as the ESI was supposed to develop an association agreement for network access. 28

We have studied this in detail elsewhere (cf. e.g. Brunekreeft, 2002, 2003), and we will summarize it here briefly.

252

Electricity Market Reform

Whereas this by and large succeeded for the ESI, this failed in the GSI, leaving the government no option but to intervene. However, as this contribution is on the ESI, we will continue with an examination of the developments in the ESI.29 High network charges are against consumers’ interests, but as long as they are within a reasonable range they are unlikely to arouse too much political attention. More important is that competition died off after a first wave of excitement. Retail competition for domestic users is problematic. Switching rates are low and third-party suppliers are in financial distress. Although consumers perhaps do not switch because they are satisfied with their incumbent supplier and although potential competition may work, we observe that active competition is not a great success. Müller and Wienken (2004) estimate that roughly 40% of the household market is effectively closed, because the margin is below cost. The developments on the wholesale market were similar and highly remarkable (see above). Undoubtedly there has been quite strong competition, which in the beginning resulted in renegotiation of old contracts by large users (industry and retailers). The presence of competition and traders acted as a threat in the bargaining game. The prices for large users, which are an indicator for wholesale prices, came down strongly, presumably squeezing out the air resulting from productivity increases made in the 1990s and which had not been passed through. As shown in Figure 8.7 above, wholesale prices at the power exchange in Leipzig were very low; as low as fuel costs and substantially below cost recovery. This short-lived success has depressed entry: the first 6 or 7 years of liberalization have not or have hardly seen third parties in generation and most planned projects were never realized. If anything, firms left the market, while the assets in the market became more concentrated through mergers and acquisitions. The low entry activity reflects different issues. The low wholesale price, policy uncertainty about next institutional steps (regulation or not) and discriminatory behavior by the network operators will all have contributed to hesitant new entry. As we will discuss in Section 8.3.3, this is now changing. Weak regulation of a strongly vertically integrated industry (and weak enforcement of unbundling) implied difficult times for competition. Complaints about discrimination against third parties have been persistent. Indeed, the first association agreement was most certainly not pro-competitive. Moreover, during the first years after market opening, the Cartel Office has been active to settle unresolved issues and pursue abusive behavior. Moreover, the institutional framework of vertical integration without effective regulation of the network charges created the incentives for a margin squeeze: in case of doubt, the integrated firms will make (excess) profits on the network, not on the commercial business. The resulting low margins were unattractive for third parties. Summing up all the points above, we conclude that among other effects, effective regulation will widen the retail and generation margin (i.e. higher wholesale prices) and make abuse of the network or system-operation more difficult. All in all, effective regulation will promote new entry in generation and retail and thereby promote new investment. 8.3.1.3. Regulation and network investment Regulation and the choice between cost based or incentive based have potentially substantial effects on network investment. Incentive based may be good for short-run efficiency but may impede long-run network investment. Recall from Section 8.3.1.2 that the difference between cost based and incentive based is not clear-cut but rather a gradual matter of accents; the same applies for the reflections below. 29

The interested reader may refer to Brunekreeft and Twelemann (2005, Section 2.2) and references quoted therein for further details on the GSI.

Energy Policy and Investment in the German Power Market

253

It is not implausible that the explicit step of creating regulation and a regulator as already carried out by the Energy Act decreases (policy or regulatory) uncertainty. This can have a stimulating effect on investment. A second effect concerns the institutional choice for the BNA, which is part of the Ministry of Economics. Independence, being one of the leading regulatory principles, is thereby violated. How this could work out depends on the interests of the Ministry. Although it does have stakes, the federal Ministry is not a major shareholder in the power industry; the Ministry’s primary interest will be the consumer. This implies that the regulator might be under political pressure to lower the network charges if this is politically opportune.30 Other effects concern the choice between cost based and incentive based regulation. Though being still ambiguous in an empirical sense, cost-based approaches are seen as inefficient and generally wasteful of resources (gold-plating) and dependent on details biased toward over-capitalization; this was one of the drivers to move away from cost-based approaches (cf. Beesley and Littlechild, 1989, p. 456). The long-run perspective reverses the argument. Gold-plating may be inefficient but might be good for investment. Gilbert and Newbery (1994) point out that, in an uncertain world, the expected deviations from the reasonable outcome are smaller under cost-based regulation than under incentivebased regulation. Importantly, this increases the credibility of the regulator to stick to previously announced policies. In other words, incentive-based regulation can impede network investment as it reduces the regulator’s credibility. Peltzman (1976) pointed out the “buffering hypothesis”, which means that rate-of-return regulation reduces the firm’s exposure to market risk as compared to no regulation. Wright et al. (2003) extend the argument for price-cap regulation. In terms of demand uncertainty, risk under the price-cap regulation is lower than without regulation, similar to rate-of-return regulation. In contrast, in terms of cost uncertainty, risk under price-cap regulation is higher than without regulation. The arguments imply that the firm’s risk-adjusted cost of capital might be higher under price-cap regulation. All else equal, this means that investment may be lower under price-cap regulation. Lastly, as Spence (1975) pointed out incentive-based regulation has poor incentives for investing in quality. A price-cap regulated firm can increase profits at the expense of quality. Regulators in the Netherlands, Norway and the UK, for instance, have adjusted the price-cap regulation for quality incentives. The new Energy Act in Germany allows the BNA to make the necessary quality adjustments. As argued in Section 8.3.1.2, what we would expect is that if long-term network investment becomes more important relative to short-term efficiency, incentive-based regulation will be modified to cost-based type of regulation and quality-adjusted regulation. It appears that this is happening in the UK where the new distribution price controls which, came into force in April 2005, included sliding scales and used-and-useful test for capital overspending. 8.3.2. The policy on RES, CHP and the CO2 emission trading scheme German environment-related policy has the following targets: ●

30

Under the EU burden-sharing agreement to implement the Kyoto climate protocol, Germany has committed itself to reducing its greenhouse gas emissions by 21% between 2008 and 2010, as compared to the 1990/1995 emission levels.

This contrasts to telecommunications, where the government was the major shareholder of Deutsche Telekom, although its stakes reduced gradually by floating the shares. Furthermore, the situation is in contrast to the state level; the states are stakeholders in the power industry, and there we might see the directions go the other way around.

254

●

●

Electricity Market Reform

CHP generation is to play an important role in achieving these targets. In accordance with the national CO2 reduction strategy the ESI committed itself to reducing CO2 emissions through an increase of CHP generation by at least 20 million tonnes by 2010, which would mean nearly a doubling of CHP generation to 20% in 2010 compared to 2002 levels. According to the CHP Act, CO2 emissions are to be reduced through an increase of CHP generation (10 million tonnes by 2005; 23 million tonnes by 2010). The red–green government aimed at doubling the share of RES by 2010 from 5% to 10%.

8.3.2.1. Support for RES and CHP RES and CHP are supported by the Renewable Energy Act (EEG) and the CHP Act, respectively. The support mechanisms for EEG and CHP plants are basically a subsidy, although different in detail. While EEG plants get a fixed remuneration depending on technology and plant size, the payment for CHP plants varies with the market price. CHP in Germany has a production share of 10%, of which 60% is gas- and 40% coal- or lignite-fueled (Fig. 8.2). Support for CHP is a plain subsidy over and above the market wholesale price. The arrangement is the result of stranded cost compensation, after it turned out that CHP became unprofitable after liberalization. The CHP Act applies only to CHP plants that were in operation when the Act entered into force and will be phased out. The Act does only apply to new plants if they replace exiting plants (modernization) and for small CHP plants below 2 MWel and fuel cells. As a result, the CHP Act does little for investment which expands CHP capacity. RES are promoted by the RES Act (EEG), which arranges a feed-in charge system with a take-off obligation: a “take and pay” system. Like the CHP support now, the pre-liberalization system used to be a predetermined subsidy on the “market price”. With liberalization the market prices and thereby the feed-in charging fell substantially, pushing the RES plants into financial distress and suppressing new investment. The government decided to change the system by fixing the feed-in charges independent of market developments. The feed-in payments are generous, with a minimum payment of €5.5 c/kWh for wind and €43.4 c/kWh for solar. The support mechanism distinguishes between technologies, vintages and sites, thereby increasing overall efficiency. The costs of the feed-in mechanism are socialized over all end users. Whilst under the old pre-liberalization mechanism, each DNO had to bear the total costs of RES in their area individually, the EEG has established a mechanism whereby the costs are spread countrywide. The DNO, to which the RES plant is connected, is obliged to take-off the energy, but passes this on to the TSO (TNO) to which it is connected. The TSOs spread the burden equally among themselves and calculate a nationwide compensation charge. They then pass it on proportionally to the suppliers in their region, who in turn pay the compensation charge and pass through the costs into the end user price. In 2004, the share of RES was about 9% and the calculated compensation charge €9 c/kWh; with a wholesale price of €3.3 c/kWh, this amounts to a “RES tax” of €0.51 c/kWh.31 A “take and pay” system of feed-in charges and take-off obligation affects competition only in an indirect fashion. The system implies that RES and conventional sources do not compete directly. Indirectly, the conventional suppliers face reduced residual demand (which is total demand minus the exogenous supply of RES), which brings the price down. Also, we should expect that as the capacity-load margin increases (excess capacity), competitive 31

Compare Haas et al. (2006) for an overview and impression of European policies and experiences.

Energy Policy and Investment in the German Power Market

255

pressure increases, which further reduces prices. Moreover, under a system of fixed feed-in charges and take-off obligation the RES do not directly compete with each other. If the share of RES is moderate this is acceptable, but if the share is substantial a large part of the market will effectively be exempted from competition. If the RES share grows, the designs of the market, network regulation and a RES support mechanism might need to be reconsidered. Network connection charges are shallow, meaning that new generation assets only pay for the cost of connecting to the first network connection point, whereas the costs of network upgrades beyond this connection point are borne by the network operator. The network operators are obliged to connect new plant as long as the request is reasonable and, if necessary, undertake network enforcements. There is a new debate, however, about an estimated €800 million for an HV network upgrade which would be necessary to facilitate offshore wind projects. Evidently, the industry argues to pass through these costs. Plants connected to the distribution network (distributed generation), including CHP and RES, receive a network charge rebate. The calculation of grid charges is based on the assumption that all electricity is fed into the high-voltage transmission grid. The payment from the DNO to the TSO, however, is based on the actual annual peak load which the DNO gets from the TSO, reduced by a coincidence factor. As a result, if there are plants connected to the distribution network, the payment which the DNO receives from grid users may exceed the charges he has to pass through to the TSO. DG receives these avoided network charges. More generally though, while the support for RES does lead to more decentralized generation, there is no explicit policy on distributed generation yet. 8.3.2.2. CO2 emission trading Emission trading started at the beginning of 2005, as part of a European Union-wide emission trading scheme. The first trading period is a pilot phase, lasting until 2007. The second trading period will last from 2008 until 2012. Each EU member state had to draw up a national allocation plan, defining the overall emission targets for the various sectors (macro plan), including the targets for those sectors covered by the emissions trading scheme (ETS) (industry, energy), and the method of allocating CO2 permits to individual plants (micro plan). The allocation of permits to individual plants is based on two principles: grandfathering based on historical values for existing plants and a kind of benchmarking for new plants. Permits are allocated to existing plants on the basis of historical emissions multiplied by a reduction factor, whereas new plants receive the permits based on their expected emissions with an upper limit set by modern coal plants and a lower limit set by CCGTs. In both cases, permits are allocated free of charge. How will the allocation plan affect investment in new power plants and thereby the environment, generation adequacy and competition? The leading principle is that irrespective of whether the CO2 permits are auctioned or are free of charge, there is an opportunity cost corresponding to the market price of the permits, pushing up marginal costs. They will thus be passed through into the electricity wholesale price. If the permits are auctioned then evidently they are real (variable) costs. If the firms receive the permits free-of-charge they earn windfall profits equal to the quantity of permits times the price. The more relevant effect is on new investment. The CO2 price as such has a merit-order effect: it makes gas less expensive compared to coal in terms of marginal costs, which may increase the load factor of gas plants and thereby reduces their average costs. For an investment decision the windfall profit translates into lower investment cost. Brunekreeft and Twelemann (2005) calculate the entry price, which is the price at which a new investment just

256

Electricity Market Reform

recovers full cost.32 Receiving a number of permits works out as lowering fixed investment costs and increasing variable (opportunity) costs.33 Lower effective investment costs make it more likely that new plants will be able to compete against existing plants. For existing machines the lower investment costs are bygones and the windfall profits can be passed through to the shareholders. For new entrants it makes a difference in the investment decision. For this reason, the free allocation of permits stimulates entry with new investment. If the initial allocation is free of charge, money flows into the system. As long as entry is possible and rewarded with new permits, this leads to excessive entry and capacity. In the long run, the profitability of plants is brought back to a normal rate of return by (inefficiently) low load factors. At least initially new entry is likely to be more efficient with lower specific CO2 emissions and thus existing plants are likely to face the lower load factor; one would anticipate the early retirement of these plants. Although auctioning the permits is superior from an efficiency point of view, allocating the permits free of charge stimulates new investment (more and sooner) which is good for competition and supply security. The effect on technology is less optimistic. As soon as allocation deviates from “best practice” (product benchmark) and instead differentiates between different technologies (technology benchmark), the technology choice is distorted. The relative advantage new RES should have under a system of auctioned CO2 permits vanishes, if the permits are allocated free of charge and according to a technology benchmark. Furthermore, if the permits are allocated according to a technology benchmark (by and large corresponding to the emission rate of state-of-the-art machines), then replacing an old inefficient, high carbon machine with a new efficient, low carbon machine implies less permits, which in turn means smaller windfalls. At the very least, this postpones the replacement. Art. 10 of the German National Allocation Act specifies a transfer rule, which addresses this problem. If an old plant is replaced by a new plant, the permits of the old plant can be transferred to the new plant for 4 years. In an insightful study, Bode et al. (2005) argue that transfer rule heavily distorts competition as it puts new entrants (who cannot “replace old machines”) at a disadvantage: for the same investment an incumbent replacing its old machine would get more CO2 permits then an entrant not replacing an old machine. Somewhat surprising though, Bode et al. (2005) also claim that (with unlimited validity of the transfer rule) the transfer rule does not speed up replacement. This counterintuitive claim seems to be due to fact that the analysis lacks an explicit dynamic factor and thus a timing problem. Explicitly including a dynamic factor and timing (e.g. demand growth or costreducing learning) repairs this point and causes the transfer rule to speed up replacement. Overall we conclude that a system allocating CO2 permits free of charge (inefficiently) supports competition with new entry and generation adequacy. The effect on the environment is less clear. Having a CO2 system at all evidently supports the environment, but technology benchmarking may well have detrimental effects. The transfer rule is good for the environment but may damage competition too much.

32 There are two key numbers, reflecting the merit-order effect. For an CO2 price of less than €30/tonne CO2 the entry price is about €52/MWh due to a low load factor. With an CO2 price of more than €30/tonne CO2 the entry price is about €36/MWh, with a high load factor. The numbers are sensitive to assumed fuel prices and efficiency levels. Compare also Pfaffenberger and Hille (2003) for similar findings. 33 An alternative way of reasoning (leading to the same result) is to argue via the revenue side of net present value. Allocating the permits free of charge does not have an effect on expenses, but it does increase market price. Hence, the system will make new investment more attractive than it otherwise would be.

Energy Policy and Investment in the German Power Market

257

8.3.3. Generation adequacy 8.3.3.1. Hands-off policy on generation adequacy The power crisis in California in 2000/2001, the power black-outs in New York, London and Italy in 2003 and many near black-outs in recent years triggered concerns about the incentives of the liberalized power markets to provide adequate capacity. The overall issue is reliability, including both generation and networks. We concentrate on generation here; an impression of the network side has been given in Section 2.1.4. Reliability in turn covers two aspects: security and adequacy. Supply security is the ability of the system to respond to short-term disturbances; this requires sufficient reserve capacity and is typically the SO’ task. Supply adequacy (or, in our case generation adequacy) reflects sufficient long-term investment in such a way that the system functions under standard conditions; this begs the question as to whether the market provides sufficient incentives to invest. This is controversial and the impression is that policy-makers in the USA are less confident in the market than in Europe. A primary problem is fluctuating and uncertain demand (and of course, the fact that electricity power cannot be stored and the fact that supply should meet demand at all times), which implies that there will be peaking units with a very low load factor. In other words, the costs of a peaking plant should be recovered in only a couple of hours. If we assume, for instance, that a peaking plant has annualized costs of €40,000/MW/year, we need 10 hours per year and a price as high €4000/MWh to recover costs. If generation units are paid only for real production, then the prices are called energy-only prices. A system of energy-only prices is typically what the spontaneous market design will be. Theory predicts that scarcity will push up prices, which will attract new investment which in turn will reduce scarcity and so on, until an equilibrium is found. This is convincing, yet there are reasons to be cautious. Individual consumers cannot (at least not in current circumstances) be shut down individually; hence consumers have weak incentives to contract for (reserve) capacity. Two points weaken this argument. First, if large consumers can be shut down individually, the total market may be sufficiently responsive; the question is what is sufficient? Second, developments with so-called smart meters, which can be used to disconnect individual households, are fast. A further argument why markets may be slow to respond to scarcity prices is these very high prices are simply unrealistic (Joskow, 2003). Joskow and Tirole (2004) point out that even such very extreme situations and subsequent extremely high prices are very sensitive to the discretionary behavior of the TSO. Furthermore, most systems have a maximum price; in many parts of the USA, bids are capped at $1000/MWh. And even if they are not capped explicitly, there is justified concern that prices higher than this might trigger government interference. As pointed out in Brunekreeft and McDaniel (2005) this may be a vicious circle ending in a low-capacity equilibrium. We see the academic controversy reflected in policy, where there is a wide variety of policy measures. In the USA, a system with capacity obligations is popular. In Europe, concern has been expressed by the European Commission with its supply security package of December 2003. Some countries in Europe have explicit policies like capacity payments (e.g. Spain) or reserve contracting (Sweden and the Netherlands34). Most countries, however, have a hands-off policy: that is, explicitly doing nothing (except perhaps monitoring) and leaving it to the market (e.g. Norway and the UK). The German approach is also hands-off although it has not been made explicit. The background is more practical; Germany has a long tradition of excess capacity on which the system still relies and the investment question is not yet urgent. Para. 51 of the Energy Act requires the monitoring of supply security by the Ministry of Economic Affairs, and in the case of installed capacity (taking account of interruptible contracts) not being adequate, para. 53 then allows the government to organize a tender for additional capacity, in line

258

Electricity Market Reform Table 8.4. Age of German generation plants (in MW). Type Hard coal Lignite Nuclear Gas Oil Other

⬎30 years

30–10 years

⬍10 years

10,635 9570 2223 7291 4879 183

17,457 6207 21,340 6980 2044 1109

768 5465 0 3293 39 1853

Source: Ziesing and Matthes, “Energiepolitik”, DIW-Wochenbericht 48/2003. 9000 8000 Million /year

7000 6000 5000 4000 Generation T&D Total Total west Total east

3000 2000 1000 2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

0

Fig. 8.10. Investment in the German ESI. Source: Data from Karl, “Ifo Schnelldienst”, various years.

with art. 7 of the EU E-Directive 2003. It is safe to conclude that generation adequacy is not a policy issue in Germany yet. 8.3.3.2. Generation capacity and investment in Germany Is this justified? Basically we observe that (as elsewhere) both investment levels and generation reserve margins have dropped in the last 5–10 years. However, in recent years, both have been restored. The reserve margin in Germany has been studied closely in Brunekreeft and Twelemann (2005) and it seems that all that has happened is that excess capacity has been reduced without endangering continuity of supply. Yet at some point new investment is required. First, to meet new demand. Second, to adjust to technological progress (certainly eyeing the environment). Third, provided that phasing out goes ahead as planned, to replace the nuclear power plants which are to be decommissioned. Fourth, to replace old and depreciated machines. Table 8.4 below suggests that a large share of the generation plants in Germany is rather old and will need to be replaced soon. As mentioned before, though, investment activity is picking up.35 Figure 8.10 shows how investment fell steeply after the all-time high following reunification. But clearly, the fall halted 34

The system in the Netherlands has been designed but the required additional reserve capacity has currently been set to zero, and hence the system is inactive at the moment. 35 See, for example, the August 2005 investment survey among 200 industry done by ZEW (www. zew.de).

Energy Policy and Investment in the German Power Market

259

and in the meantime and investment is increasing again. Further, the increased wholesale prices make new projects attractive and indeed attract new entry. There are projects by companies like Statkraft, yet the most interesting development comes from mainly municipal distributors/retailers joining forces and investing in new generation plants. Examples are Trianel and SüdWestStrom. A reason which is sometimes heard, is to reduce the dependence on the big producers, from which we may conclude that competition has not been working all that well. A problem for new entrants has been the availability of sites for power plants. While completely new sites are difficult to find and often meet the resistance of the local public, existing sites are difficult to get hold of because they are in most cases controlled by the incumbents. For example, there was interest from municipal utilities in the south to build a CCGT plant on the Obrigheim site, a decommissioned nuclear plant. Yet EnBW who owns the site, refused to make it available for such a project, putting forward grid-related arguments. Lastly, the regulator and regulation should be expected to ease new entrants’ lives and as explained the CO2 ETS as well as the RES policy appears to support new investment. All in all, the German ESI may need new investment but it is likely to come. 8.3.3.3. New capacity: gas or coal? For several years after liberalization, gas-fueled CCGT was seen as basically the only option for new plants. In 1991, the European Commission lifted restrictions on the use of gas for electricity production. Gas prices were relatively low. With CCGT, new gas-fueled technology was highly efficient. The relatively low capital costs and short construction times and life duration made new gas an attractive investment in the liberalized market. However, coal and lignite are on the rise again, at least in terms of announcements and expectations. Even EnBW in the south of Germany ponders the possibility of building new coal plants, although not that long ago transportation costs in this region, which is far from both domestic and imported coal, were thought to be prohibitive. CCGT plants still benefit from high fuel efficiency and low capital costs, but the high gas price has turned against it. Gas projects have also suffered from the lack of gas network regulation and problems with third-party access, a situation that can be expected to be improved by the new Energy Regulator. This particular problem was presumably worsened by the controversially approved E.On-Ruhrgas merger, in as far as new entry was expected to be with gas-fueled plant. The European Commission is worried about a high European gas import dependence (from northern-Africa, Russia and the Middle-East). Alternatives are for instance to rely more strongly on liquefied natural gas (LNG) and more on indigenous sources like coal. On the other hand, as becomes clear from Figure 8.9 hard coal prices are increasing as well. Heavily increased demand from especially China has increased upward pressure on the coal price. There are also increasing costs of transportation, apparently especially due to the Chinese claim on shipping of steel. It is sometimes expected that the high coal price will not last, due, for instance, to exploitation of new mines and transportation capacity, while gas prices are expected to remain high. The CO2 ETS makes coal relatively more expensive compared to gas.36 However, the CO2 price must be rather high to have a significant effect. On the other hand, if fuel efficiency increases, ceteris paribus CO2 emission per MWh goes down. Although the same holds for 36 However, coal does not contain methane (CH4) which puts coal at a relative advantage if methane should be part of an emission scheme.

260

Electricity Market Reform

Table 8.5. Power plant projects in Germany. Project name/ location

Company

Plant type/ fuel

Size in MW

Comments

Hamm

Trianel

Gas CCGT

800

Under construction. 28 municipal utilities from Germany, Austria and the Netherlands have shares in this project, to go on-online in September 2007.

HuerthKnapsack

Statkraft

Gas CCGT

800

Project was taken over from intergen. Under construction.

Lubmin

Concord Power

Gas CCGT

1200

Irsching/ Ingolstadt

Eon/N-Ergie/ Mainova

Gas CCGT

800

Herdecke

Mark-E/Statkraft

Gas CCGT

400

Lingen

RWE

Gas CCGT

800–1000

Boxberg

Vattenfall Europe

Lignite

700

Additional capacity, not replacing old lignite plants.

Neurath

RWE

Lignite

1100 or 2200

Replacing old lignite units.

Karlsruhe/ Heilbronn

EnBW

Hard coal and/or gas

SüdWest-Strom

Hard coal or gas

400–800

No final decision on fuel yet, coal would have relatively high transport costs, municipalitybased company.

Trianel

Hard coal

700–800

As with the Trianel CCGT project, municipal utilities can buy shares in this project. Replacing a 300 MW coal plant.

No final decision on fuel yet, coal would have relatively high transport costs.

Datteln

Eon

Hard coal

1000

Hamm

RWE

Hard coal

2 * 750

Hamburg

Vattenfall Europe

Hard coal

700

Source: Company information, various sources.

CCGT, rather strong technological advances are expected in more efficient coal and lignite plants (supercritical plant and clean-coal technology) (cf. e.g. Bode et al., 2005). While both hard coal and gas can be bought on the market and are potential options for new entrants, lignite is a different game. As it is too expensive to transport lignite over long distances, due to its low energy density, lignite plants are generally located right next to the mine, the mines are owned by the generators and the fuel is shoveled from the mine into the plant without going through any form of market. Consequently, the marginal costs of lignite plants are anyone’s guess. For new entrants, there is no way they can get access to lignite and new lignite plants will be built by the incumbents. Currently, it is mainly RWE that is about to replace its old plants by new ones. We conclude that the future of coal as a fuel for the German ESI looks brighter than is sometimes thought. Looking at projects under construction and announced projects from mid-2005, gas and coal have about the same share (see Table 8.5).

Energy Policy and Investment in the German Power Market

261

8.4. Concluding Remarks This contribution examines energy policy in Germany. The primary focus is on the effect of various policies, which directly or indirectly relate to the energy market, on investment in the ESI in Germany. Investment in turn affects competition, environment and supply adequacy. The policies we examine are threefold. First, we study the policy related to liberalization and regulation of the ESI. On July 13, 2005 a new Energy Act entered into force implementing the EU Directive of 2003. The key point of the new Energy Act is to remove negotiated TPA and instead establish regulated TPA. It can be concluded that the previous system, which did not have effective regulation did not work. Network charges are high both in international comparison and relative to the end user price. Competitive margins are low, which impedes effective competition. The new system installs a sector-specific regulator (BNA) and regulation. The regulation is as yet cost based, but the Act explicitly allows the option to switch to ex-ante, incentive-based regulation. In practice this means a shift of emphasis toward ex-ante, forward-looking capping of revenues for a predetermined period and presumably a stronger reliance on benchmarking of different firms. As set out extensively in this contribution, we expect that the regulation of network access will strongly support the development of competition in both generation and retail; we already observe that investment activity in generation assets is starting to take off. On the other hand, it can be argued for a variety of reasons that the shift toward incentivebased regulation, which aims at short-run efficiency, tends to have detrimental effects on (long-run) network investment. This can be defended though, because currently the networks are viewed as being in good shape albeit inefficient and the Energy Act does allow quality-adjusted regulation. Second, Germany has a strong tradition of supporting the environment. The policies on RES, CHP and the CO2 emission trading are dominating the debate at the moment. The support schemes and network connection arrangements for RES and CHP, although with different background, are generous and should be expected to support further new investment. The costs of the schemes are passed through to end users and network charges, respectively. Examination reveals that the numbers are too small to make RES and CHP responsible for the recent increase of end user prices. The CO2 price is surprisingly high and as the German power production relies on coal the CO2 mark up in Germany is high as well. Exactly this seems to explain much of the recent price increase. The system of allocating the CO2 permits free of charge, whilst inefficient, stimulates new investment and thereby promotes competition and supply adequacy. Oddly, as a consequence of having a technology benchmark, the new investment need not be in clean technology. Third, despite controversial debate on generation adequacy elsewhere, Germany has no explicit policy on generation adequacy. Leaving the theoretical question as to whether the energy-only market will provide sufficient capacity aside, we observe that capacity margins and investment levels have dropped in the last 6 or 7 years, but have been restored recently. Moreover, generation assets in Germany are old and replacement and modernization are required soon. At the same time, we observe that investment activity (at least announced) is definitely picking up again. Challenging the conventional wisdom that gas will dominate the future, it seems that hard coal has a brighter future than sometimes thought.

Acknowledgment The authors would like to thank Paul Joskow, Wolfgang Pfaffenberger and Perry Sioshansi for useful comments and remarks.

262

Electricity Market Reform

References Beesley, M.E. and Littlechild, S.C. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20(3), 454–472. Bergman, L., Brunekreeft, G., Doyle, C., von der Fehr, N.-H., Newbery, D.M., Pollitt, M. and Régibeau, P. (1999). A European Market for Electricity? Monitoring European Deregulation 2. CEPR/SNS, London/ Stockholm. BMWA (2005). EWI/Prognos-Studie: Die Entwicklung der Energiemärkte bis zum Jahr 2030. BMWA Dokumentation Nr. 545, May 2005, Berlin. Bode, S., Hübl, L., Schaffner, J. and Twelemann, S. (2005). Oekologische und wettbewerbliche Wirkungen der Uebertragungs- und der Kompensationsregel des Zuteilungsgesetzes 2007 auf die Stromerzeugung. HWWAReport 252. Brunekreeft, G. (2002). Regulation and third party discrimination in the German electricity supply industry. European Journal of Law and Economics, 13(2), 203–220. Brunekreeft, G. (2003). Regulation and Competition Policy in the Electricity Market; Economic Analysis and German Experience. Nomos, Baden-Baden. Brunekreeft, G. (2004). Regulatory threat in vertically related markets: the case of German electricity. European Journal of Law and Economics, 17(2), 285–305. Brunekreeft, G. and McDaniel, T.M. (2005). Policy uncertainty and supply adequacy in electric power markets. Oxford Review of Economic Policy, 21(1), 111–127. Brunekreeft, G. and Twelemann, S. (2005). Regulation, competition and investment in the German electricity market: RegTP or REGTP. Energy Journal, Special issue, 99–126. Bundeskartellamt, (2001). Bericht der Arbeitsgruppe Netznutzung Strom der Kartellbehörden des Bundes und der Länder. 19 April 2001, Bundeskartellamt, Bonn. Clingendael Institute (2004). Study on the Energy Supply Security and Geopolitics. Report prepared for DG TREN, January 2004, The Hague. EC (2005). 4th benchmarking Report. European Commission, DG TREN, Brussels. Gilbert, R.J. and Newbery, D.M. (1994). The dynamic efficiency of regulatory constitutions. Rand Journal of Economics, 25(4), 538–554. Grout, P.A. and Zalewska, A. (2003). Do regulatory changes affect market risk? LIFE Working Paper 03-022, LIFE, Maastricht University, Maastricht. Growitsch, C. and Müsgens, F. (2005). The economics of restructuring the German electricity sector. Mimeo, University of Cologne, Cologne. Growitsch, C. and Wein, T. (2004). Negotiated third party access – an industrial organization perspective. Working Paper No. 312, University of Lueneburg. Haas, R., Glachant, J.-M., Keseric, N. and Perez, Y. (2006). Perspectives for long term competition in the continental European electricity market. In F.P. Sioshansi and W. Pfaffenberger (eds.), Electricity Market Reform: An International Perspective, Elseviers Scientific, 2006 (forthcoming). Joskow, P.L. (1974). Inflation and environmental concern: structural change in the process of public utility price regulation. Journal of Law and Economics, 17(1), 291–327. Joskow, P.L. (1989). Regulatory failure, regulatory reform, and structural change in the electric power industry. Brookings Papers on Economic Activity; Microeconomics, Vol. 1989, 125–208. Joskow, P.L. (2003). The difficult transition to competitive electricity markets in the U.S. MIT, Boston, Mass, May 2003. Joskow, P.L. (2005a). Transmission policy in the United States. Utilities Policy, 13(2), 95–115. Joskow, P.L. (2005b). Incentive regulation in theory and practice: electricity distribution and transmission networks. Mimeo draft August 8, 2005, MIT, Boston. Joskow, P.L. and Tirole, J. (2004). Reliability and competitive electricity markets. Working Paper CMI EP 53, University of Cambridge, Cambridge. Müller, Chr. and Wienken, W. (2004). Measuring the degree of economic opening in the German electricity market. Utilities Policy, 12(4), 283–290. Müsgens, F. (2004). Market power in the German wholesale electricity market. EWI Working Paper, No. 04.03, EWI Cologne.

Energy Policy and Investment in the German Power Market

263

Peltzman, S. (1976). Toward a more general theory of regulation. Journal of Law and Economics, 19(2), 211–240. Pfaffenberger, W. and Hille, M. (2003). Zukünftige Energieoptionen: Sicherung der Investitionen in die Elektrizitätsversorgung. Report, Bremer Energie Institut, Bremen. Spence, A.M. (1975). Monopoly, quality and regulation. Bell Journal of Economics, Vol. 16, 417–429. Wade (World Alliance of Decentralised Energy) (2005). World Survey of Decentralised Energy 2005, Edinburgh. Wright, S., Mason, R. and Miles, D. (2003). A Study into Certain Aspects of the Cost of Capital for Regulated Utilities in the U.K. Report, February 13, 2003, Smithers & Co, London. Wuppertal Institut (Hg.) (2004). Braunkohle-ein subventionsfreier Energieträger? Kurzstudie. Wuppertal.

This page intentionally left blank

Chapter 9 Competition in the Continental European Electricity Market: Despair or Work in Progress? REINHARD HAAS,1 JEAN-MICHEL GLACHANT,2 NENAD KESERIC1 AND YANNICK PEREZ3 1

Institute of Power Systems and Energy Economics, Vienna University of Technology, Vienna, Austria; Head of the Department of Economics, University Paris XI, Cedex, France; 3Department of Economics, University Paris-Sud 11, Cedex, France

2

Summary This chapter examines the perspectives for competitive electricity markets in Continental Europe. In most Continental European countries restructuring of the electricity market started in the late 1990s and is still going on. The object of this chapter is to investigate past developments in this market and to analyze which conditions are necessary to enhance competition in the long run. Currently, the major obstacle for one common European electricity market is a general lack of competition in virtually all local and national wholesale as well as retail electricity markets because the number of competitors is too low, or because barriers to entry and incentives to collude remain too high. Our major conclusion is that several conditions are necessary to bring about effective competition in the Continental European electricity market: (i) a complete separation of ownership of the transmission grid and the generation and supply in all countries and submarkets; (ii) sufficient transmission capacity for creating a larger market; (iii) adequate margins in generation capacity; (iv) a sufficiently large number of generators to share this capacity; (v) a secure and competitive supply with primary fuels (notably natural gas). As it is not likely that these conditions will be fulfilled the prospects for a vibrant competition in Continental Europe are bleak.

9.1. Introduction The restructuring of electricity markets in most Continental European (CE) countries started in the late 1990s, and is still going on. This process was triggered by the European Commission (EC) directive, 1996, “Directive for a common electricity market”. The major 265

266

Electricity Market Reform European electricity sub-markets

2

1

6

5

1. UK & Ireland 2. Nordiccountries 3. Iberian peninsula 4. Italy 5. Western Europe 6. Eastern Europe 7. South-Eastern Europe

7

3 4

Fig. 9.1. Electricity sub-markets in Europe in 2005.

motivation for this directive was the EC’s conviction that liberalization, price deregulation and privatization would directly lead to competition in generating, as well as supply which would then result in lower prices for the whole of Europe. The EC’s main expectation in the directive was the belief that “market forces (would) produce a better allocation of resources and greater effectiveness in the supply of services”.1 In June 1996, after years of discussion, the European Council of Energy Ministers reached an agreement with the European Parliament on a market liberalization directive, and 6 months later passed the full directive concerning common rules for the internal market in electricity which, with the intention of restructuring the European power industry, became law in February 1997. The initial intention of the EC was the creation of a common European electricity market, but this area still consists of at least seven different sub-markets which are separated by insufficient transmission capacities, and differences in conditions for access to the grid (Fig. 9.1). Furthermore “the evidence from Europe and the USA suggests that there are a number of conditions for successfully liberalizing electricity markets.” (Newbery, 2002; Glachant and Finon, 2003). We have identified conditions which could bring about a common competitive European electricity market, thus leading to competitive electricity prices. They are: ●

●

●

1

access to the grid, this requires unbundling of generation from transmission, and of supply from distribution; supply adequacy, adequate capacity in generation and transmission as well as access to primary energy sources (e.g. natural gas); market structure, ownership and number of generators and suppliers;

EC Communication Services of general interest in Europe, OJ C 281, 26. September 1996, p. 3.

Competition in the Continental European Electricity Market

● ● ●

267

design of the market place, notably the ease of entry for new players; regulatory governance; environmental issues, which are playing an increasingly important role.

The goal of this chapter is to analyze the evolution of the European electricity markets and to discuss future developments with respect to competition (see former treatments in Bergman et al., 1999; de Paoli, 2001; Glachant and Finon, 2003; Jamasb and Politt, 2005; Glachant and Lévêque, 2005; as well as the special issue of the Energy Journal, 2005). This chapter covers most of what is currently called “Continental Europe” (CE): Austria, Belgium, Czech Republic, France, Germany, Hungary, Poland, Portugal, Slovenia, Slovakia, Spain and Switzerland. It is organized as follows: Section 9.2 provides the historical context with major data and facts for the liberalization of the CE markets. Section 9.3 describes EC and national governments’ market liberalization initiatives. Section 9.4 presents major changes country by country. Section 9.5 discusses the evolution of the markets corresponding to government initiatives. Section 9.6 describes the market’s remaining problems; followed by conclusions in Section 9.7.

9.2. Background: Facts, Figures and History Before 1990, almost every electricity supply industry in Continental Europe was vertically integrated with a captive franchise market, either state owned (the majority of cases) or under price-regulated mixed private/public ownership (as in Belgium, Germany and Switzerland) (see Chapter 1). Regulated area monopolies prevailed in all countries. Until the end of the 1990s, the standard model was “an effectively vertically integrated franchise monopoly under either public ownership or cost-of-service regulation” (Newbery, 2005). Although electricity networks were typically synchronized over wide areas, interconnections of areas under different transmission system operators (TSOs) were frequently guided by security rather than economic considerations. However, most trade in the past was due to economic benefits of arbitrage during peak load hours. Real electricity liberalization in Europe started with Britain’s restructuring and privatization in 1990, demonstrating that vertical unbundling and the creation of wholesale electricity markets was actually feasible (see Newbery, 2005). Jamasb and Pollitt (2005) argue that the centralized approach to market liberalization because of European Electricity Directives has succeeded in maintaining the pace of reform in the original EU-15, and in a number of associated and accession countries, and, as well as achieving a certain degree of standardization of structures, institutions and rules in national markets. However, the problems created by initially concentrated market structures have been reinforced by a wave of subsequent mergers, and the low level of interconnection that reduces the scope for fostering competition by imports (Glachant and Lévêque, 2005). Yet, ownership structures and degree of vertical integration were quite different among the following countries: ●

● ●

In France, Italy, Portugal, the former Czech-Slovak Republic, Poland, Hungary and Slovenia a strong state-owned vertically integrated monopoly dominated the ESI. This centralized structure typically led to a single dominant player, such as Electricité de France. In Spain and Switzerland, vertical integration was strong, but with a handful of companies. In Germany there were about 10 generators integrated with transmission, but only partially integrated with supply.

268

●

●

●

●

Electricity Market Reform

In Austria there was one large generator which was integrated with transmission, and about 14 regional suppliers fully integrated with distribution. In the Netherlands there was an upward vertical integration with the distribution companies controlling the grid and the generators. In Belgium, the large majority of the power sector has been private for decades. The private generator Electrabel is supervized and controlled by the mother company, Tractebel. Belgium, Germany, Spain and Switzerland were the only countries in the mid-1990s where private ownership among generators prevailed (tempered in Germany and Switzerland by the local public ownership of distribution and supply, and the former “State enterprise” nature of Endesa in Spain). It contrasted with the state-owned enterprises in France, Italy, Portugal and the remaining Central and Eastern countries.

9.2.1. Development of demand and supply About 2300 TWh were consumed in the CE area in 2004. The largest electricity markets are currently in Germany, France, Italy and Spain. Highest per capita demand was in Luxemburg, Belgium and Switzerland. The lowest per capita demand was in Poland, Hungary, Portugal and Slovakia. Demand growth per year was strongest in Spain (⫹5.0%), Portugal (⫹4.9%), and Austria (⫹3.1%). In Poland and Germany demand increased by about only 1%/year. In the whole of the CE, electricity consumption grew from 1% to 3% per year between 1999 and 2004. Details are depicted in Figures 9.2–9.4.

Poland The Netherlands (145 TWh) (111 TWh) Germany NL (554 TWh) Czech Republic Belgium (61 TWh) Slovakia (88 TWh) Austria CZ (26 TWh) (52 TWh) France Hungary CH Switzerland (445 TWh) SI (38 TWh) Slovenia (65 TWh) (12 TWh) Spain (235 TWh)

Italy (322 TWh)

Portugal (46 TWh)

Fig. 9.2. Electricity consumption in CE countries in 2004.

269

Competition in the Continental European Electricity Market

In 2004, generation in CE countries was mainly from fossil thermal power plants (mainly coal) with 51%, followed by nuclear (34%) and hydro (12%). Other renewable (mainly wind) have contributed 3%. As shown in Figure 9.5 the distribution of generation sources across CE countries is rather uneven. In most countries thermal power dominates, in Italy and the Netherlands with more than 80%, in Poland almost 100%. In France, Belgium and Slovakia nuclear power plays the most important role. Only in Austria and Switzerland does hydro power prevail. Specific electricity consumption CH ES SK SI PT PL NL LU IT HU DE FR CZ BE AT 0

2000

4000

6000 8000 kWh/cap

10000

12000

14000

Fig. 9.3. Comparison of electricity consumption per capita in CE countries in 2004. Demand increase (2000–2004) CH ES SL SK PO PL NL LU IT HU DE FR CZ BE AT 0.0

1.0

2.0 3.0 Average (%/year)

4.0

Fig. 9.4. Growth of electricity demand in CE countries (average 2000–2004).

5.0

270

Electricity Market Reform Generation mix of continental european countries (TWh in 2004) 100 90 80 70

%

60 50 40 30 20 10 0 AT

BE

FR

DE Hydro

IT

LU

Nuclear

NL

PO

Thermal

ES

CZ

HU

Renewables

PL

SK

SL

CH

Others

Fig. 9.5. Comparison of the fuel mix for generation in CE countries in 2004. Source: UCTE (2005).

9.2.2. Generation capacity and load Capacity margin is different among countries as can be seen from Figure 9.6 . However not all gross capacity are available for generation. This is especially true for hydro capacity (Austria, Spain) and old fossil plants (Italy). For example, Italy, Austria and the Netherlands which are net importers of energy also exhibit such an apparent excess capacity margin. Figure 9.7 depicts the evolution of generation capacity over the last 10 years in CE. The growth in capacity is mainly from wind power and fossil fuel power plants. 9.2.3. Development of imports and exports In 2004, the total amount of electricity exchanges between CE countries stood at about 300 TWh. This is equal to about 13% of consumption and is frequently limited by constrained cross-border transmission capacity. Figure 9.9 show the physical2 electricity exchanges between CE countries. France is the biggest net exporter among CE countries with net exports of almost 67 TWh, followed by Czech Republic and Poland. The major importing countries are Italy with 51 TWh followed by The Netherlands and Hungary with 17 TWh and 7 TWh, respectively. The percentage of imports and exports of total electricity demand in CE countries is depicted in Figure 9.8. Smaller countries like Switzerland, the Czech Republic, Slovakia and 2

To some extent these flows are not due to contracts between countries but just because of loop flows (e.g. from Germany to Poland to Czech Republic back to Germany).

271

Competition in the Continental European Electricity Market Installed gross capacity in CE countries versus peak load in 2004 130 Others Renewables Thermal Nuclear Hydro Available capacity Peak load 2004 (MW)

120 110 100 90 GW

80 70 60 50 40 30 20 10 0 AT

BE

FR

DE

IT

LU

NL

PO

SE

CZ

HU

PL

SK

SL

CH

Fig. 9.6. Installed gross and net generation capacity (except auto producers) and maximum load in CE countries 2004.

Historical development of gross generation capacities in CE 600,000 550,000 500,000 450,000 400,000

MW

350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 1995

1996 Hydro

1997

1998

Nuclear

1999

2000

Thermal conv.

2001

2002

Renewables

2003 Others

Fig. 9.7. Evolution of generation capacity in CE 1995–2004. Source: UCTE, national reports.

2004

272

Electricity Market Reform Percentage of imports/exports CH ES SL SK PO

Imports

Exports

PL NL LU

⫺108%

IT HU DE FR CZ BE AT

⫺60

⫺40

⫺20

0

20

40

60

Percent of total consumption Fig. 9.8. Imports/exports as percentage of electricity demand in CE countries. Source: UCTE (2005)

Luxemburg, with around 20% of net exports related to domestic consumption, transfer the largest shares of their electricity to and from other countries. Figure 9.9 shows the evolution of imports and exports since the start of liberalization. There is a trend toward a slight increase, but no dramatic boost. 9.2.4. Past and current issues of transmission The bulk of the transmission and distribution network in Europe was built between 1950 and 1990 prior to the introduction of market liberalization and has had few additions in recent years. Figure 9.10 shows the highest percentages of net transfer capacity (NTC) used in 2004 between CE countries.3 Due to the operating complexity of the European meshed network, commercial capacity and physical capacity differ. Hence, the interconnection capacity is defined by European transmission system operator (ETSO) as “NTC”. The most congested lines are between Italy and its neighboring countries; and between Spain and Portugal. But the borders between Germany, Austria and the Czech Republic are next. In principle the congested lines need a special mechanism so as to be managed in an economic way (see Section 9.5.6). The existing CE network was built to guarantee a good level of technical reliability and to give some room for managing peak load problems. Now, it is supposed, it is to be used in a more economic way, under optimization processes of scarce capacity, and to produce price convergence in a single European market perspective. 3

Further details are documented in Table 9A.3 in the Annex. It shows current cross-border transmission capacity (Net Transfer Capacity, NTC), as published by the European Transmission System Operators (ETSO) for winter 2004/2005, physical flows 2004 and maximal possible (theoretical) annual energy flows between the countries.

273

Competition in the Continental European Electricity Market Electricity net imports in CE (1998–2004) 100 Austria

90 Spain

Spain

Luxembourg Austria Austria Luxembourg Portugal Spain Luxembourg Luxembourg Luxembourg Hungary Hungary Portugal Hungary Portugal Spain Belgium Belgium Hungary Belgium Hungary Luxembourg Belgium Spain Belgium Belgium Netherlands Netherlands Luxembourg Netherlands Netherlands Belgium Netherlands Netherlands

80 70 60 TWh

Austria Spain

50

Netherlands

40 30 Italy

Italy

Italy

Italy

Italy

Italy

Italy

1998

1999

2000

2001

2002

2003

2004

20 10 0

Austria Spain Luxembourg Portugal Switzerland Slovenia Hungary Germany Belgium The Netherlands Italy

Electricity net exports in CE (1998–2004) 110 Switzerland

100 90 Austria

80

TWh

70 60

Switzerland Austria Poland Switzerland Czeh.Rep. Germany Poland Czeh.Rep.

Austria Switzerland Poland

Switzerland Germany Poland Germany Poland

Czeh.Rep.

Switzerland Switzerland Germany Germany

Poland

Poland Czeh.Rep. Czeh.Rep.

Czeh.Rep.

Czeh.Rep.

Austria

50

Switzerland Slovenia

40 30

Slovakia France

France

France

France

France

France

France

Germany Belgium

20

Poland Czeh.Rep.

10

France

0 1998

1999

2000

2001

2002

2003

2004

Fig. 9.9. Evolution of net imports and net exports over the period 1998–2004. Source: UCTE (2005).

274

Electricity Market Reform Maximum percent use of transmission capacity SLO–IT ES–PO SK–HU CZ–DE AT–IT CH–IT FR–DE FR–IT CZ–AT DE–AT PL–CZ 0

50

100 (%/year)

150

200

Fig. 9.10. Major bottlenecks in CE transmission grid measured as percentage of use of transmission capacity per year in 2004. Source: UCTE (2005) (for details see Table 9A.3 in the Annex).

Basically, in the new competitive system, European interconnectors have to allocate electricity flows from low-cost regions to high-cost regions, and by doing so, they are expected to produce both a price convergence and a redistribution of stakeholders’ welfare. 9.3. How the System Changed: Political Issues Of Restructuring The restructuring of the CE electricity market was triggered by the EU Directive on ‘Common Rules for the Internal Market in Electricity’ which came into force in February 1999.4 The major intention was to create a common European electricity market, EC (1997).The major issues of this Directive (officially named 96/92) were: ● ● ●

Minimal requirements for the unbundling of generation and transmission. Minimal market opening, expressed by the consumption size of “eligible customers”. Different approaches for the access to the grid (negotiated or regulated, TPA or single buyer).

However, each national government within the EU had to “transpose” the EU Directive into national law and national rules. An overview on the major milestones is provided in Table 9.1. In practice, the major area of action within the European liberalization project was “providing access to the market”. Far less attention was paid to the issues of restructuring generation & supply and designing market places as well as ensuring adequate generation and transmission capacity. Independent energy regulators were introduced in all countries

4

As already mentioned in some countries in CE (Germany, Poland and Spain) steps towards liberalization were set already before the EU Directive went into force.

Competition in the Continental European Electricity Market

275

Table 9.1. Milestones of reforming in Continental Europe. 1996

EU-15

February 1997

EU-15

1998 1998 1998 February 1999

Spain Poland Germany EU-15

2001 2001

Austria EU-15

2003

EU-25

2003 2004

Spain EU-15 ⫹ 10

2004

EU-25

2005

Portugal, The Netherlands EU-25

2007

European Council of Energy Ministers and Parliament reached agreement on a market liberalization directive This “Directive concerning common rules for the internal market in electricity” (Directive 96/92/EC) became valid while waiting up to two more years for its transposition by countries Introduction of a Spanish centralized pool Introduction of TPA (market opening: 22%) 100% market opening in one step Directive went into force after a 2 years transposition delay: market opening due the directive in Austria, Belgium, France, Italy, Spain, Portugal and the Netherlands between 30% and 35% 100% market opening (in a second step) Approval of the “Directive of the European Parliament and the Council on the promotion of electricity from renewable energy sources in the internal electricity market (RES-E Directive)” (European Parliament and Council, 2001 – Directive 2001/77/EC) Approval of the “Directive concerning common rules for the internal market in electricity” (officially Directive 2003/54; usually named “the Second Directive”) 100% market opening Extension of the EU to 25 member countries, new CE member countries to open their market with 30% minimum Electricity Directive 2003/54 due to be transposed by member states All non-domestic customers made eligible in the EU in July 2004 An EU Regulation on cross-border electricity trade came into effect (Regulation 1228/2003) in July 2004 100% market opening Due to Electricity Directive 2003/54, 100% market opening in all EU-25 countries in July 2007

except Germany (and Switzerland which is not part of the EU). In addition, environmental issues were also treated very prominently. On the contrary, aside from a minimal unbundling, the restructuring of utilities and the design of market places was not tackled comprehensively by governments in most countries (few exceptions: Spain created a centralized pool, and Italy divested generation capacities). The milestones of reforming the electricity sector in Continental Europe are given in Table 9.1. 9.3.1. Providing non-discriminatory access to the market and to the grid The first important requirement for a competitive electricity market is non-discriminatory access to the grid. Therefore, a prerequisite for competition is the unbundling of generation

276

Electricity Market Reform

Table 9.2. Types of unbundling of transmission system operators and access to the grid in CE (as of December 31, 2004a. Country

Unbundling TSOb

TSO

Ownership

Access to the grid 2004

Austria

Legal (APG); Management (TIWAG, VKW)

APG (90%), TIWAG (6%), VKW (4%)

100% public, 100% public, 51% public

rTPA

Belgium

Legal (2005: ownership)

ELIA

100% Electrabel (2005: floated)

rTPA

Czech Republic

Legal

CEPS

51% CEZ, 49% public

rTPA

France

Legal

RTE

100% EdF

rTPA

Germany

Legal

RWE Netz, E-ON-Net, EnBW-Net, Vattenfall Transmission

100% RWE 100% E.ON 100% EnBW 100% Vattenfall Europe

nTPA

Hungary

Legal

MAVIR

100% public,

rTPA

Italy

Ownership

GRTN

100% public

rTPA … eligible customers SB(rTPA) – captive customers

Luxembourg

Management

ELIA (BE) RWE-Netz (DE)

100% ELIA 100% RWE

rTPA

The Netherlands

Ownership

TenneT

100% public

rTPA

Poland

Legal

PSE (Polskie Sieci Elektroenergetyczne S.A.)

100% public

rTPA

Portugal

Ownership

REN

100% public

rTPA … eligible customers SB(rTPA) … captive customers

100% public

rTPA

100% public

rTPA

Slovenia

Ownership

ELES

Slovakia

Legal

SEPS

Spain

Ownership

REE

Switzerland

No

Regional vertically integrated companies

a

rTPA No

rTPA: regulated third-party access; nTPA: negotiated third-party access; SB: single buyer model. Legal: legal separation of transmission and generation. Source: CEC, 2005; CEC, 2004, company reports, Power in Europe, own investigations. b

Competition in the Continental European Electricity Market

277

Table 9.3. Electricity directive implementation in CE countries. Country Austria Belgium Czech Republic France Germany Hungary Italy Luxembourg The Netherlands Poland Portugal Slovenia Slovakia Spain Switzerland

Market opening (%) (February 19, 1999)

Market opening (%) (January 1, 2005)

30 35 0 30 100 0 30 0 33 22 30 0 0 45 0

100 90* 55 68 100 67 79 57 100 80 100 75 66 100 0

Eligible customers (January 1, 2005) All >10 GWh* >0.1 GWh All non-households All All non-households >0.1 GWh >20 GWh All All non-households All All non-households All non-households All No final customers

*Figures for Wallonia. Full market opening in Flanders region. Source: CEC (2001); CEC (2005).

and supply from transmission. This means that access to transmission and distribution should be offered to all market participants at reasonable and non-discriminatory prices. So far, the experiences with respect to unbundling between generation and transmission in CE have been different. In Belgium, Spain, Portugal and Italy unbundling of generation and transmission by ownership was achieved either by full independence of the transmission company or by the flotation of a transmission subsidiary. In other countries, especially in Germany and France, only legal unbundling took place. In Switzerland, so far unbundling has only been done by means of internal management measures. These give no structural guarantees for avoiding discrimination in access to the grid, in particular when no independent regulator is able to monitor the behavior of grid managers. Table 9.2 provides the current status of unbundling. The second issue is the regime of access to the grid. Table 9.2 shows access to the transmission grid in various Western European countries (CEC, 2005). Access to the grid has been regulated in all countries except Germany where it was introduced in June 2005. The third issue is market opening. The geographically, and timely different opening of the markets led, at least to some distortions regarding free choice of supplier. Table 9.3 and Figure 9.11 depict the opening of the market in different CE countries between 1999 and 2005. Some countries like Germany, the Netherlands, Spain, Portugal and Austria have legally fully opened their markets, while others, like France, Luxemburg and Czech Republic have only partially opened their’s. In Switzerland (which is not member of the EU) there is currently no competition in supply.

9.3.2. The new institutional and regulatory environment In all countries, except Germany and Switzerland, independent regulatory authorities have been founded. An overview of these regulatory authorities and their staff and budget is

278

Electricity Market Reform Market opening 100 2005 2003 2001 1999

90

70 60 50 No market opening

Percent market opening

80

40 30 20 10 0 DE AT ES NL PO BE PL

IT

SL FR HU SK LU CZ CH

Fig. 9.11. Market opening in CE as of January 1, 2005. Source: CEC (2005) and earlier benchmarking reports.

given in Table 9.4.5 Powers vary widely from one country to another, but common core tasks are: ● ● ●

to ensure that unbundling is achieved; to regulate access to the grid; to regulate tariffs for the use of the transmission and distribution grid.

In practice, the current European regulatory governance consists of a decentralized framework on national levels in an incomplete process of convergence. Countries have6 established nationally based regulatory authorities which are administered by nationals. Access to the national TSO’s grid and operating system is regulated nationally. All this is done legally and with recourse to courts, while the European Directives and Regulations provide only a broad common frame. However the EC or the European Court of Justice can block this or that excess on a case-by-case basis. For example, in summer 2005 the European Court deemed illegal the “grandfathering” priority given to incumbent Foreign suppliers of the Dutch grid interconnections. 9.3.3. The promotion of renewables Currently, the promotion of electricity from renewable energy sources (RES-E) plays an important role in the energy policy of the EU. The major policy reasons are: (i) reducing the

5

It would be interesting to analyze whether there is any correlation to the size or budget of the regulator and the working of the market. Yet, unfortunately, such an analysis would go beyond the scope of this chapter. 6 Except Germany and Switzerland.

279

Competition in the Continental European Electricity Market Table 9.4. Budget and staff of regulatory authorities in CE. Country

Name (year of foundation)

Austria Belgium Czech Republic Germany France Hungary Italy Luxembourg The Netherlands Poland Portugal Slovenia Slovakia Spain Switzerland

E-Control (2001) CREG ERU (Bundesnetzagentur, 2005) CREG HEO AEEG ILR DTE URE ERSE Energy Agency URSO CNE No

Budget 2004 (Million euro)

Staff 2004

8.3 11.3 3.8 – 13.8 6.2 18 0.7 5.1 7.7 7 1.5 1.5 20.7 No

66 74 88 (180 in 2005) 108 95 100 32 55 267 51 22 57 175 No

Origin of budget P L P – E P P P P P L/P P P P No

L:levy on operators; P: public budge; No: does not exist. Source: European regulators, AIE, CEC (2004); Kaderjak (2005). Table 9.5. Renewable electricity targets as share of electricity consumption in the EU-25 member states. Country Austria Belgium Czech Republic France Germany Hungary Italy Luxembourg The Netherlands Poland Portugal Slovak Republic Slovenia Spain Switzerland

RES-E penetration 1997 (%)

RES-E target for 2010 (%)

70 1 4 15 4.5 0.7 16 2.1 3 1 38 18 30 20 68

78 6 8 21 13 3.6 25 5.7 9 7 39 31 34 30 No

dependence on energy imports; (ii) reduction of greenhouse gas emissions. To meet this target the EU has defined ambitious objectives which have been formalized in the “Directive of the European Parliament and the Council on the promotion of electricity from renewable energy sources in the internal electricity market (RES-E Directive)” (EC 2000). According to this directive, RES-E generation should reach a total share of 22% of electric production in 2010 from a level of 12% in 1998 (EC, 2000). Table 9.5 specifies the indicative targets for the share of RES-E for every CE country to be met by 2010.

Electricity generation (TWh/year)

280

Electricity Market Reform

60 50 40 Biogas Solid biomass Bio waste Geoth Wind onshore Wind offshore Hydro small scale Hydro

30 20 10 0 AT BE CZ FR DE HU IT LU NL PL PT ES SK SI CH

Fig. 9.12. Electricity generation from renewables in 2003 by country. Source: EEG TU Wien, GREEN-X database.

Figure 9.12 depicts the amounts of various RES-E technologies, country by country. Hydropower is the dominant source, but ‘new’ RES-E’s such as biomass and wind are starting to play a role. Wind energy has had a yearly growth rate of about 35% per year over the last decade. Biomass is especially popular in countries like Poland, where it is commonly cofired in existing coal power plants to meet the negotiated renewable energy share. Yet, the higher costs of RES-E technologies, compared to existing conventional power plants, advocate financial support. As the choice of instruments has not been prescribed or harmonized in Europe, every country has adopted its own. In Table 9.6 an overview is provided of promotion schemes for RES-E in EU-15 countries for the Year 2004. Feed-in tariffs are currently used in most of the CE countries. This instrument has so far turned out to be the most effective for a fast deployment of significant shares of RES-E.7 The promotion of wind energy has so far been the most successful in this context. As Figure 9.13 depicts, that due to feed-in tariffs in Germany and Spain, considerable capacity of wind power was constructed up until 2004.

9.4. Comparison of Developments by Country The developments toward competition in the countries and sub-markets have been quite different so far, as can be seen in Table 9.7. Germany started with a 100% market opening without any restructuring of the industry. Later on, a rapid merger process took place, resulting in the disappearance of half the generating, transmission companies. Moreover, the German idea of competition was unique because no regulatory authority was created. It soon became evident that high grid charges, discrimination with respect to access to the distribution network, and high transaction costs of the negotiated TPA were major problems for this model, in particular, because of the hundreds of regional or local distribution grid companies. Finally, in 2005 a regulatory body was created. 7

For a comprehensive comparison of the relative efficiency of guaranteed feed-in tariffs, bidding system and exchangeable quotas systems, see Finon and Perez (2006).

Table 9.6. Overview of the main policies for the promotion of RES-E in CE countries (as of end of 2004).{not in reference} Country

Main electricity support schemes

Comments

Austria

Feed-in tariffs (presently terminated) combined with regional investment incentives

Feed-in tariffs have been guaranteed for 13 years. The instrument was only effective for new installations with permission until December 2004. The active period of the system has not been extended nor has the instrument been replaced by an alternative one.

Belgium

Quota obligation system/TGC combined with minimum prices for electricity from RES

Federal government has set minimum prices for electricity from RES. Flanders and Wallonia have introduced a quota obligation system (based on TGCs) with obligation on electricity suppliers. In Brussels no support scheme has been implemented yet. Wind offshore is supported on the federal level.

France

Feed-in tariffs

Germany

Feed-in tariffs

For power plants ⬍12 MW feed-in tariffs are guaranteed for 15 years or 20 years (hydro and PV). For power plants ⬎12 MW a tendering scheme is in place. Feed-in tariffs are guaranteed for 20 years (Renewable Energy Act). Furthermore soft loans and tax incentives are available.

Italy

Quota obligation system/TGC

Obligation (based on TGCs) on electricity suppliers. Certificates are only issued for new RES-E capacity during the first 8 years of operation.

Luxembourg

Feed-in tariffs

Feed-in tariffs guaranteed for 10 years (for PV for 20 years). Also investment incentives available.

The Netherlands

Feed-in tariffs

Feed-in tariffs guaranteed for 10 years. Fiscal incentives for investments in RES are available. The energy tax exemption on electricity from RES was finished on January 1, 2005.

Portugal

Feed-in tariffs combined with investment incentives.

Investment incentives up to 40%.

Spain

Feed-in tariffs

Electricity producers can choose between a fixed feed-in tariff or a premium on top of the conventional electricity price; both are available during the whole life time of the RES power plant. Soft loans, tax incentives and regional investment incentives are available.

Czech Republic

Feed-in tariffs (since 2002), supported by investment grants revision and improvement of the tariffs in February 2005.

Relatively high feed-in tariffs with 15 year guaranteed duration of support. Producer can choose between fixed feed-in tariff or premium tariff (green bonus). For biomass cogeneration only green bonus applies.

Hungary

Feed in tariff (since January 2003) combined with purchase obligation and tenders for grants

Medium tariffs (6–6.8 c/kWh) but no differentiation among technologies. Actions to support RES are not coordinated, and political support varies. All this results in high investment risks and low penetration.

Poland

Green power purchase obligation with targets specified until 2010. In addition renewable exempted from the (small) excise tax

No penalties defined and lack of target enforcement.

Slovak Republic

Program supporting RES and EE, including feed-in tariffs and tax incentives

Very little support for renewable. Main support program runs from 2000, but no certainty on time frame or tariffs. Low support, lack of funding and lack of longer-term certainty make investors very reluctant.

Slovenia

Attractive feed-in system combined with long-term guaranteed contracts, CO2 taxation and public funds for environmental investments

None

Switzerland

Source: Huber et al. (2005).

282

Electricity Market Reform Installed wind capacity 2004 SL HU SK CZ LU CH PL BE FR PT AT NL IT ES DE 0

2,000 4,000

6,000 8,000 10,000 12,000 14,000 16,000 18,000 MW

Fig. 9.13. Wind capacity in CE by the end of 2004. Source: EWEA.

In Austria the market was legally opened in two steps: 33% in 1999 and 100% in 2001. In 2001 a voluntary spot market (EXAA) was founded. Since 2000 a discussion has been ongoing concerning several models of national and cross-border mergers and takeovers. Yet, so far only minority shares of some suppliers have been sold to the French EdF, or the German EnBW and RWE. In France, more than 90% of capacity is concentrated in EdF, with two potential competitors who have been institutionally linked to it. These links have been weakened in order to make them independent in the near future, and have been opened to new entrants, notably Electrabel and ENEL. These “fringe generators” are CNR, a hydro generator, and SNET, a subsidiary of Charbonnages de France which produces 8.5 TWh by dispatchable coal plants. The transmission business was made a subsidiary in the second half of 2005, and could be floated as soon as 2006. EdF, itself will put around 20% of its shares on the market before the end of 2005. The major feature, in the Czech Republic and the Slovak Republic, of restructuring was the break-up of the former vertically integrated public utility into generation, grid and supply companies. Furthermore, in the meantime, parts of the generation and supply companies have been privatized. In 1993, the Czech Republic spread about 31% of CEZ shares among investors (individuals and funds). As an attractive offer was not received for the rest of CEZ, further privatization has been delayed so far. In the Slovak Republic, 66% of the generator SE is being privatized (2005). In Hungary, Slovenia and Italy steps were taken to reduce the power of the former generation monopoly. Currently, however, it appears that in these countries the former monopolists still have a strong position in the market (ENEL kept 50% of the Italian generation capacity, plus the cash made by selling the rest of its plants – as “Gencos” or by the sale of transmission and distribution grid shares). In the Netherlands, until 1998, generation was dominated by four large regional companies: EPZ, EPON, UNA and EZH, who jointly owned the generator Sistema Electríco Publico (SEP). The Dutch government’s initial idea was to combine liberalization in supply with the concentration in generation by merging the four companies and SEP. This attempt should have

283

Competition in the Continental European Electricity Market Table 9.7. Differences in reforming and market design in various countries. Process of market opening

Mandatory pool

Voluntary day ahead exchange

Futures market

Privatization process

Divestment of generation capacity

Takeover, merger within the country

AT

Fast (2 years)

No

Yes (EXAA)

Yes (EEX)

Moderate

No

Under discussion

BE

Slow

No

No

No

*

No

No

CZ

Moderate

No

Yes (2004)

No

No

No

No

DE

Very fast

No

Yes

Yes

*

No

Yes, half electricity generation plus Ruhrgas

FR

Slow

No

Yes

No

No

No

Yes, two fringe generators

HU

Moderate

No

No

No

Moderate

No

No

IT

Slow

No

Yes (since 2004)

No

Yes

Yes

Yes, mainly abroad (ENEL in SK)

LU

Slow

No

No

No

N.A.

No

No

NL

Moderate

No

Yes (APX)

No

Yes

No

Yes, mainly from abroad

PL

Fast

No

PO

Moderate

SK

Moderate

No

No

No

Yes

No

No

SL

Moderate

No

Yes (2003)

No

Moderate

Moderate

No

ES

Moderate

Yes

No

No

*

No

No

CH

No

No

No

Yes (EEX)

*

No

No

Yes

No

Moderate

Yes

Moderate

No, but intended with Spain

No

No

Yes, moderate

Moderate abroad

*Major generators were already largely private before liberalization started.

created a “national champion” that would be able to compete on the European scene (van Damme, 2005). Yet, the merger failed because these companies could not agree. The major restructuring feature was then the sell out of half of the former largely public-owned generator to companies from abroad (Electrabel, Reliant, E.ON). Another trend is the vertical reintegration of generators and suppliers for example, by the purchase of power plants by suppliers. After a series of mergers and takeovers, two large

284

Electricity Market Reform

Dutch companies survived and are now integrated into generation, distribution and supply (ESSENT and NUON). The TSO TEnnET and its subsidiary, the power exchange (PX) of Amsterdam, have been 100% state owned for some years. In Belgium, the process has been dominated by the incumbent company Electrabel, which is controlled by the Suez group (France) through the intermediate engineering contractor Tractebel. A “second” Electrabel was developed outside Belgium by collecting 15,000 MWe plant capacity, mainly in Europe (the Netherlands, Poland, Hungary, Italy, France, Spain). In spring 1999, Tractebel pretended to become a liberalization champion. They split their companies into parts while keeping control over all of them all. In 2005 however, Electrabel and Tractebel were merged to increase their stock market size. They understood that being one of the oligopolistic players on the European and worldwide market was more profitable than to stay linked to the limited Belgian market (Verbruggen and Vanderstappen, 1999). The Spanish approach initially looked like being one of the most ambitious. However, the structure of the industry with two dominant producers integrated in distribution and supply was never changed. As a result, after the introduction of a centralized pool8 in 1998, the issue of market power exerted by the two largest incumbent generators was very soon raised. Crampes and Fabra state in (2005): “The 1997 reform did not succeed in introducing effective competition but retained an opaque regulation which has been subject to continuous governmental interventionism. …” Note, that due to scarce interconnection capacity between Spain and neighboring countries foreign utilities have not been very influential in the Spanish pool so far. The issue of market power is still – in 2005 – the major problem in Spain and could be reinforced by the take over attempt of the first gas company Gas Natural of the first electrician Endesa. In 2005, an investigation in the competitiveness of the Spanish market was conducted by Perez-Arriaga, and the new government is reviewing the rules with the view to changing them. As well as transmission, there were only four significant companies, all largely private and vertically integrated. While the former government blocked the merger of the two largest utilities (Endesa and Iberdrola), it allowed the takeover of Hidrocantabrico by EDP. Furthermore, when Endesa put 5% of its activities up for sale, it was bought by ENEL of Italy, as Endesa had just taken control of Elettrogen in Italy (Soares, 2003). In Portugal, the hard process of the privatization of the EDP, and the creation of a competitive affiliate (Sistema Electrico Nao-Vinculado, SENV) has been shaping the reform process so far. The idea was to split the national electricity system into two sub-systems: the public utility system SEP and the independent system SENV. SEP and SENV are not generator s, but sub-systems of the national electrical system. The former has to satisfy demand under the principal of a uniform tariff on the mainland, which moderates the application of market rules. It also has centralized planning. The latter has no responsibility for public service and comprises of two sub-systems: the non-binding system (SENV) and the independent producers. The SENV operates according to market rules and comprises of producers, distributers and eligible customers. Non-binding producers and customers are allowed to use the public utility system grid for a fee (Soares, 2003). Other objectives, since the start of reform in Portugal, have been to create a “national champion” by merging gas and electricity monopolies (which was refused by the European Competition Authority) and a joint Iberian market with Spain (The MIBEL project). Yet, so far this Mibel has been repeatedly postponed and currently, it is being planned to put it into practice in 2006. One problem is that “without substantial enhancements to interconnection, 8

While the participation in this pool is in fact mandatory, market participants are also allowed to enter into physical bilateral contracts (Crampes and Fabra, 2005).

Competition in the Continental European Electricity Market

285

it should be clear that the impact of the Spanish market on the highly concentrated, Portuguese market can only be marginal, and the impact of the Portuguese wholesale market on the Spanish minimal” (PiE 437, p. 3). With respect to divestment of capacity, Italy was the only country in Continental Europe where the former state-owned champion had been privatized and had to give away generation capacity (Lorenzoni, 2003). Currently, however, ENEL is in a comfortable position because it is still the largest electricity producer in a market with congested borders and a congested internal grid and can act as a private company with the cash generated by its divestiture. ENEL has now a market share of 50% of generation capacity, and an Italian PX has been opened. In Switzerland a draft law providing for ultimately complete opening of the Swiss electricity market was rejected by the Swiss population in a referendum in 2002. Another draft law providing for market opening for larger industrial customers was provided for discussion in 2004. Given the legislative procedure and a possible new referendum, first steps of market opening can be expected, in the case that the law is finally approved, at the earliest in 2008 (CEC, 2005). Eastern European countries are physically integrated within the western European grid, and have taken the first steps toward adopting the “western model” with regulated third-party access for larger customers. There has been partial privatization of companies within the industry (except in Slovenia) and the reduction of barriers to international trade. But, like the rest of Europe, each reform is unfinished in regard to its market design and the existing market power of the dominant player. The typical Eastern European market structure is made up of a dominant wholesaler and a competitive fringe. The competitive fringe is strongly limited by long-term contract structures that often allow the dominant wholesaler to deploy the generators, so being able to deny other companies’ access to surplus capacity that has not been contracted in advance (Kaderjak, 2005). It is also the case concerning the support for renewable energy which often takes the form of a feed-in tariff under which the power is sold to the dominant wholesaler, thus consolidating its position even more. Poland and Hungary were the forerunners of reform in Eastern Europe. Poland introduced TPA in 1998, and the Czech Republic and Hungary conducted unbundling of generation and transmission in the early 1990s. Hungary established a regulator in 1994 and started privatization of supply and most of generation in 1995. At the same time the gradual removal of price subsidies was started (Kaderjak, 2005).

9.5. The Markets: Structures and Performances The markets’ structures and performances after the start of liberalization can be measured in different ways. However, the evolution of electricity prices is, presumably, the most important indicator. A desirable outcome of a single European electricity market is the achievement of a lower price and a price convergence through wholesale and retail competition (Jamasb and Politt, 2005). Hence, in this section, after having examined the characteristics of the markets and the markets’ structures, focus will be put on prices changes for differing groups of customers and in various regions. 9.5.1. Characteristics of the markets Table 9.8 depicts the markets in CE. In particular, the degree of liquidity in spot markets and bilateral markets is indicated. As the European Commission states (CEC, 2005): “Ideally spot

286

Electricity Market Reform Table 9.8. Trading in CE in 2003/2004: spot markets, centralized versus mandatory pools and OTC.

Total supply 2004 (TWh) AT BE CZ DE FR HU IT NL PL PO SK SL ES CH

51.8 87.5 61.4 445.1 554 38.2 322 6.3 110.9 144.8 45.5 26 12.3 234.5

Spot market/ centralized mandatory or voluntary pool EXAA No OTE EEX (Sp.m.) Powernext (Sp. M.) No Yes (2005) APX (Sp. M.) Pol-PX No No SLOex OMEL (C.p.) No

Volume (TWh 2004) 1 0.3 39 7.5 2 15 1.1

0.36 204

OTC (TWh 2004) N.A. N.A. N.A. 342 300 N.A. 56 240 N.A. N.A. N.A. N.A. 5 N.A.

Source: CEC (2005) and own investigations.

markets should have enough liquidity to give a reliable and transparent price signal (… .). The normal benchmark from other commodity markets is that the volume of trade (of long term contracts) should be roughly 10 times the amount of physical delivery.”9 As Table 9.8 shows, no CE market is approaching this level. As can be seen from Chapter 1 of this book no CE country is in the list of the 12 most competitive countries. Most of the CE countries have few and relatively illiquid organized markets (PXs) for electricity. Such spot markets which exist, as in Poland and Slovenia, trade less than 5% of the total electricity consumption. Bilateral contracts are the most frequent form of “trade” arrangement in new EU members’ states. 9.5.2. Mergers, takeovers and market concentration The industrial reference model for electricity completely changed between 1995 and 2001. It has shifted from a preference for vertical disintegration between generation, trading, and sales to final consumers toward a preference for vertical reintegration of production, trading and final sales. Among the best illustrations of the changing “industrial paradigm” are the shifting attitudes of financial markets, financial analysts, rating agencies and banks vis-à-vis disintegrated structures, especially concerning “pure” trading and “pure” generation as in merchant plants. Bankers and financiers have finally joined force with stockholders and managers of firms operating in competitive energy markets, and concluded that vertical integration is the best protection against volatility and the cyclical nature of markets. 9

In the EU, there are differences regarding the mutual role of bilateral trade (with or without use of a broker) and PXs. We cannot guarantee the data in Table 9.8 which is mainly based on an EC report (CEC, 2005). The OTC figures are likely to be higher. There is no real transparency in the markets outside PXs and therefore it is not clear what conclusions can be drawn.

287

Competition in the Continental European Electricity Market Electricity sector national and cross-border M&As in the EU 60 50

National M&As Cross-border M&As 10

40 30

10

14 13

42

20 3

10 0

3 5

10

1998

1999

22

2000

27 19

2001

2002

2003

Fig. 9.14. Number of mergers within European electricity companies from 1998 to 2003. Source: Codognet et al. (2005).

Hence, for effective competition, a large number of companies is required. This has been clearly proven by the English and Welsh examples, where the number of generators has been increased several times by the regulatory authority (as well as by investors, notably the regional distribution and supply companies, the RECs). The “merger-mania” within the CE after the start of liberalization indicates that the major strategy of the bigger incumbent utilities is competing by merging so as to purchase market shares. Figure 9.14 depicts the mergers within the EU. These activities reached a maximum number in 2004, 5 years after liberalization started. In addition to vertical reintegration, we also observed intense activity in horizontal mergers and acquisitions. The most significant example is doubtlessly Germany, where the 10 biggest electrical and gas concerns that existed at the time the European directive was adopted in 1996 have now become four. As in the German example, integration and concentration between electricity and gas is another defining feature of this new “consolidation” phase in Europe’s energy industry. Among the seven biggest electricity firms in Europe, Vattenfall and EDF have proven themselves to be anomalies because they are notably less involved in gas to date. Finally, while gas wholesale markets and concerns have persisted in courting the entry of large European and North-American petroleum and gas companies, electricity wholesale markets and electricity and gas retail markets, have not experienced any comparable influx. Thus, the upshot is a net “consolidation” of the industry on the pan-European scale, with an increasingly concentrated small number of international European firms in the sector, sometimes mockingly called the “seven brothers” in a transparent reference to the “seven sisters” of the international petroleum industry in the 20th century. Nonetheless, on a country-by-country basis, the European Union often comes across as juxtaposing domestic markets of monopolies or duopolies with a small competitive fringe in which one, two or three fringe new entrants operate. In many Eastern European countries, national companies have been sold to strategic investors from abroad, with EdF E.On, RWE, Electrabel and Vattenfall particularly active. In reaction, some countries like Czech Republic, Slovakia and Slovenia have been concerned with the retention of national champions. These national champions have the size to survive

288

Electricity Market Reform

among the larger European groups with their unfortunate consequences for the level of competition within their national market and the European competitive game. The vested interests of the dominant incumbents in this region are encouraging them to fight against greater competition which is being pushed by further reforms. How should these mergers be seen in the light of competition? In principle, mergers and acquisitions should not be a major preoccupation. On the one hand, this issue is “old hat” in European competition policy, and, on the other hand, it is an excellent lever for directly obtaining structural remedies on a European scale that would be otherwise unattainable. If, nonetheless, certain “real” problems emerge, this more likely reflects on the deficiency of certain national rulings, especially when governments or “ordinary” judges can deliberately ignore the anticompetitive effects of their decisions. This would result at the very least, in a lack of harmonization between national decisions and those taken at the European level. The E.ONRuhrgas merger in Germany which created the biggest gas and electricity concern in the western world, will remain a bone of contention and a source of confusion for a long time. However, we cannot see any simple workable solution given the unwillingness of national governments to remedy the situation. The recent strengthening of the harmonization and cooperation between national and European authorities affects only the competition authorities, and not the other national third parties that possess other real decision-making powers. See how Portugal, and more recently Spain paid for having national gas and electricity mergers. With respect to market shares in CE, in 1998, 10 generators owned 60% of the generation capacities, in 2002 it was only six (see Codognet et al., 2005). Thomas (2003) suspects that finally European-wide only “seven brothers” will remain as large generators. Of particular concern, with respect to competition, is the situation in Central Europe (France, Germany, the Benelux countries and Austria). The concentration process in the electricity generation market was especially fulminous in Germany. Mez (2003) provides an impressing and detailed description of this process. A different but converging picture is described in Finon (2003). He portrays how a dominant player like EdF in France can benefit from liberalization by exerting market power in the home market, while at the same time is pursuing an aggressive acquisition policy abroad. Verbruggen and Vanderstappen (1999) show the same for Electrabel – Distrigas group in Belgium. As can be seen from Figure 9.15, of the 13 largest generators which existed in 1999 – the year liberalization started – in CE 5 years later only nine remained. Now in Continental Europe seven large concerns dominate the market: EdF-EnBW, RWE, E.ON, Vattenfall, Endesa, ENEL and Electrabel (Haas et al., 2002). Another interesting fact is (Table 9.9) that in the ranking of the largest generators public ownership still prevails. Of special interest is that the larger European groups put special focus on extension of their interest spheres to regions which are adjacent or separated by low transmission capacity from their home area (see Fig. 9.16). Table 9.10 depicts the current market structure in CE countries. In most countries market structure is highly problematic particularly when the national grid is poorly connected with adjacent markets. It is of specific interest that potential imports vary considerably. The small countries Luxemburg, Slovakia, Slovenia, Austria and Hungary have a potential of more than 70%. In the large countries Spain, France and Italy the potential is less than 20%. 9.5.3. Wholesale electricity price evolution How electricity prices developed after restructuring is of special interest. Figure 9.17 depicts the price evolution in CE in 1999–2004. With the exception of Italy in 2004 there was some

289

Competition in the Continental European Electricity Market Number of large generators in CE VEW

Largest CE generators 1999

VEAG EnBW CEZ Iberdrola

Iberdrola

Endesa

CEZ

Elektrabel

Endesa

Bayernwerk

Electrabel Vattenfall Europe ENEL

Vattenfall Preussen Elektra RWE

Largest CE generators 2005

Major mergers and acquisitions

RWE

ENEL

E-ON

EdF

EdF/EnBW 0

100

200 (TWh)

300

400

0

100

200

300

400

500

600

(TWh) 13

9!

Fig. 9.15. Largest European electricity generators in 1999 and 2005. Source: Own investigations.

convergence of wholesale electricity spot market prices. Moreover, while volatility in 2002 and 2003 was rather high it became moderate during 2004. In the first half of 2005, prices in Western markets increased, while prices in Poland remained stable. From Figure 9.17 we derived the following effects: (i) in Western Europe, prices increases were relative to the frame and timing of liberalization; (ii) the price level is highest in areas where capacity margin is smaller, and cross-border transmission capacity is congested (Italy, the Netherlands); (iii) prices have been highest in years when there was low hydro or low nuclear availability; (iv) however, wholesale prices are increasing and are converging at the top in markets which are connected by sufficient transmission capacity. Therefore, a major question is, are these prices a result of competition? That is to say, do these prices reflect the marginal costs of the generation set or are they influenced by some kind of market power. As for example, Muesgens (2004) shows from 2001 to 2003 in Germany, the difference between wholes